Free POST Trial Brief - District Court of Delaware - Delaware

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 1 of 68

IN THE UNITED STATES DISTRICT COURT FOR THE DISTRICT OF DELAWARE TAKEDA PHARMACEUTICAL COMPANY, LTD., and TAP PHARMACEUTICAL PRODUCTS INC., Plaintiffs and Counterclaim-Defendants, v. TEVA PHARMACEUTICALS USA, INC., and TEVA PHARMACEUTICAL INDUSTRIES LTD., Defendants and Counterclaim-Plaintiffs. ) ) ) ) ) ) ) ) ) ) ) )

C.A. No. 06-033 (SLR)

PLAINTIFFS' ANSWERING POST-TRIAL BRIEF MORRIS, NICHOLS, ARSHT & TUNNELL LLP Jack B. Blumenfeld (#1014) Mary B. Graham (#2256) Rodger D. Smith II (#3778) James W. Parrett, Jr. (#4292) 1201 N. Market Street P.O. Box 1347 Wilmington, DE 19801 (302) 658-9200 OF COUNSEL: Eric J. Lobenfeld Tedd W. Van Buskirk Arlene L. Chow Dillon Kim HOGAN & HARTSON LLP 875 Third Avenue New York, New York 10022 (212) 918-3000 Philippe Y. Riesen HOGAN & HARTSON LLP Shinjuku Center Building, 46th Floor 25-1 Nishi-Shinjuku 1-chome Shinjuku, Tokyo 163-0646 Japan (81) 3-5908-4070 Attorneys for Takeda Pharmaceutical Company Ltd., and TAP Pharmaceutical Products Inc.

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 2 of 68

Richard de Bodo Lawrence J. McClure HOGAN & HARTSON LLP 1999 Avenue of the Stars, Suite 1400 Los Angeles, CA 90067 (310) 785-4600 Attorneys for Takeda Pharmaceutical Company Limited William F. Cavanaugh Stuart E. Pollack Chad J. Peterman Melissa Mandrgoc PATTERSON BELKNAP WEBB & TYLER LLP 1133 Avenue of the Americas New York, New York 10036 (212) 336-2000 Attorneys for TAP Pharmaceutical Products Inc. January 10, 2008

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 3 of 68

TABLE OF CONTENTS I. II. A. B. C. PRELIMINARY STATEMENT ......................................................................................... 1 BACKGROUND TO ACID-RELATED GASTROINTESTINAL DISORDERS AND PROTON PUMP INHIBITORS ................................................................................ 4 Peptic Ulcer Disease and Gastroesophageal Reflux Disease (GERD) .......................... 4 Treatment for Peptic Ulcer Disease and Gastroesophageal Reflux Disease .................. 4 Medicinal Chemists in 1984 Knew Little About How to Develop Lansoprazole or Other Successful PPIs ............................................................................................. 5

III. LANSOPRAZOLE WAS A DIFFICULT AND SURPRISING DISCOVERY.................... 6 A. B. 1. 2. C. 1. 2. 3. 4. D. 1. 2. 3. E. F. Small Changes Affect Not Only Activity, but also Toxicity and ADME...................... 6 A Team of Takeda Scientists Devoted Over Two Years and Synthesized Over 400 Compounds Before Discovering Lansoprazole ..................................................... 7 Takeda Scientists Synthesized a Lead Compound: Timoprazole ........................... 7 The Inventors' Efforts to Modify Timoprazole ..................................................... 9 Four Unique Hurdles Make PPI Development Unpredictable .................................... 12 Hurdle 1: The PPI must accumulate solely in the parietal cell ............................. 12 Hurdle 2: The PPI must be stable in the blood (i.e., at neutral pH) ...................... 13 Hurdle 3: The PPI must convert to an active species in the body......................... 13 Hurdle 4: The PPI active species must fit the proton pump ................................. 14 Prior Art Would Have Led Medicinal Chemists Away from Developing Lansoprazole............................................................................................................. 14 The `431 Astra Patent......................................................................................... 14 The `409 Byk Gulden Patent .............................................................................. 16 The `518 Byk Gulden Patent .............................................................................. 17 Medicinal Chemists Would Not Have Selected Trifluoroethoxy as a Substituent in 1984...................................................................................................................... 18 Teva's Reliance on DTX-703 is Baseless .................................................................. 19

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 4 of 68

IV. SUBSTANTIAL DIFFERENCES, SCOPE AND CONTENT OF THE PRIOR ART, LEVEL OF ORDINARY SKILL, AND OBJECTIVE INDICIA SHOW THAT LANSOPRAZOLE WAS NONOBVIOUS ............................................................ 20 A. B. C. D. E. 1. 2. 3. 4. V. Teva Fails to Describe a Reasonable Expectation of Success..................................... 21 Teva's Purported "Obvious Path to Lansoprazole" Uses Hindsight and Ignores the Prior Art's Teaching Away.................................................................................. 23 The Federal Circuit and this Court Have Rejected Similar Obviousness Theories Based on Similar Facts ............................................................................... 25 The Person of Ordinary Skill in the Art ..................................................................... 26 Teva Ignores the Objective Indicia of Nonobviousness ............................................. 27 Lansoprazole Exhibited Unexpected Results ...................................................... 28 Many Other Companies Tried and Failed to Produce a PPI................................. 29 Prevacid®'s (lansoprazole) Commercial Success ................................................ 29 There was an Unmet Need For Lansoprazole...................................................... 31

THERE IS NO EVIDENCE THAT THE INVENTORS OR ANYONE ELSE WITHHELD MATERIAL INFORMATION WITH AN INTENT TO DECEIVE THE PTO.......................................................................................................................... 32 A. B. C. 1. 2. 3. 4. Different Animal Models Screen Compounds For Distinct Properties ....................... 32 The IMAU Model Provides Advantages Over Other Models; It Mimics the Location, Shape, and Histology of Human Ulcers ..................................................... 33 The Inventors Did Not Intend to Deceive the Examiner by Providing IMAU Data .......................................................................................................................... 34 The Inventors and Takeda Management Relied on the IMAU Values Disclosed in the `098 Patent to Select Lansoprazole ........................................... 34 The Inventors and Their Assistants Testified to Their Good Faith in Submitting Data From the IMAU Model in the `098 Patent ................................ 36 The Raw Data Submitted in the `098 Patent's Disclosure Is Consistent With Raw Data Reported in Takeda's Own Internal Reports .............................. 37 The Inventors' Non-Attendance at Trial is Not Evidence of Intent to Deceive the PTO ................................................................................................ 38

ii

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 5 of 68

5. D. 1.

Takeda's Regulatory Affairs Department's Selection of Tests to Submit to the FDA in its IND is Not Evidence of Intent to Deceive the PTO ...................... 38 The Allegedly Withheld Data Is Not Material ........................................................... 40 The Patent Asserts that Its New Compounds Have "about 1.5-20 times or more" Superior Anti-Ulcer Effect, and Does Not Assert the 20-Fold Difference in Any Model that Teva Alleges........................................................ 40 Teva's Contentions Concerning Dog and In Vitro Antisecretory Testing Ignore the Patent's Emphasis on Anti-ulcer Effect.............................................. 41 a. b. Dog antisecretory data is irrelevant to the `098 patent......................................... 41 In vitro antisecretory data is irrelevant to the `098 patent.................................... 42 Other Models Confirm Lansoprazole's Superiority, and are Consistent With the Patent's Statements .............................................................................. 42 There is No Other IMAU Data that is Materially Different from the Data Reported in the `098 Patent ................................................................................ 43 A Reasonable Examiner Would Not Rely On The IMAU Results in Evaluating the Prima Facie Obviousness Case ................................................... 46 Teva Provides No Evidence that the Inventors Knew of the Alleged Materiality of Antisecretory Models to the `098 Patent................................................................ 48 The Record Also Contradicts Teva's Other Inequitable Conduct Allegations At Trial, Which It Now Ignores in Its Brief.................................................................... 48 Teva's Conclusory Allegations Cannot Comprise Clear and Convincing Evidence of Intent, Materiality, or the Inventors' Knowledge of Materiality, and Leave Little to No Evidence to Balance .............................................................. 49

2.

3. 4. 5. E. F. G.

VI. `321 PATENT CLAIM 2 WAS A NONOBVIOUS INNOVATION THAT PROVIDED A VIABLE COMMERCIAL LANSOPRAZOLE PRODUCT ...................... 53 A. B. C. The Problem in the Prior Art and the Solution........................................................... 53 The Difficulty in Formulating Drugs In Solid Oral Dosage Forms............................. 53 Teva Does Not Provide Any Motivation To Combine Its References, or Explain How They Were to be Combined ................................................................. 54

VII. CONCLUSION ................................................................................................................ 58

iii

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 6 of 68

TABLE OF AUTHORITIES CASES Advanced Display Sys., Inc. v. Kent State Univ., 212 F.3d 1272 (Fed. Cir. 2000) ........................................................................................... 29 Aventis Pharma Deutschland v. Lupin, Ltd., 499 F.3d 1293 (Fed. Cir. 2007) ........................................................................................... 26 Bayer AG v. Dr. Reddy's Labs., Ltd., No. 04-179-SLR, 2007 WL 3120794 (D. Del. Oct. 25, 2007)...................................21, 26, 50 Boehringer Ingelheim Vetmedica, Inc. v. Schering-Plough Corp., 106 F. Supp. 2d 667 (D.N.J. 2000)...................................................................................... 38 Bogosian v. Woloohojian Realty Corp., 323 F.3d 55 (1st Cir. 2003) ................................................................................................. 38 Budde v. Harley-Davidson, Inc., 250 F.3d 1369 (Fed. Cir. 2001) ........................................................................................... 21 CFMT, Inc. v. Yieldup Int'l Corp., 349 F.3d 1333 (Fed. Cir. 2003) ........................................................................................... 21 Cargill, Inc. v. Canbra Foods, Ltd., 476 F.3d 1359 (Fed. Cir. 2007) ........................................................................................... 51 Catalina Lighting, Inc. v. Lamps Plus, Inc., 295 F.3d 1277 (Fed. Cir. 2002) ........................................................................................... 50 Cordis Corp. v. Medtronic AVE, Inc., F.3d, 2008 WL 60499 (Fed. Cir. Jan. 7, 2008) .................................................................... 58 Demaco Corp. v. F. Von Lansdorff Licensing Ltd., 851 F.2d 1387 (Fed. Cir. 1988) ........................................................................................... 30 Dura Auto. Sys. of Indiana, Inc. v. CTS Corp., 285 F.3d 609 (7th Cir. 2002)..........................................................................................46, 49 Ecolochem, Inc. v. S. Cal Edison Co., 337 F.3d 1361 (Fed. Cir. 2000) ........................................................................................... 23 Eisai Co. v. Teva Pharmaceuticals USA, Inc., 472 F. Supp. 2d 493 (S.D.N.Y. 2006) ............................................................................49, 52 Eisai Co. v. Teva Pharms. USA, Inc., 2007 WL 1437834 (S.D.N.Y. May 14, 2007)...................................................................... 52

iv

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 7 of 68

Eli Lilly and Co. v. Zenith Goldline Pharm., Inc., 471 F.3d 1369 (Fed. Cir. 2006) ........................................................................................... 21 FMC Corp v. Manitowoc Co., 835 F.2d 1411 (Fed. Cir. 1987) ........................................................................................... 49 Governali v. Am. Tempering, Inc., No. 85-7305, 1988 WL 3843 (E.D. Pa. Jan. 20, 1988), aff'd, 856 F.2d 83 (3d Cir. 1988) .................................................................................................................................. 38 Halliburton Co. v. Schlumberger Tech. Corp., 925 F.2d 1435 (Fed. Cir. 1991) ........................................................................................... 50 Hoffman-La Roche, Inc. v. Promega Corp., 323 F.3d 1354 (Fed. Cir. 2003) ......................................................................................49, 51 Knoll Pharm. Co. v. Teva Pharms. USA, Inc., 367 F.3d 1381 (Fed. Cir. 2004) ........................................................................................... 28 KSR Int'l Co. v. Teleflex Inc., 127 S. Ct. 1727 (2007).............................................................................................21, 23, 58 Leviton Mfg. Co. v. Universal Sec. Instruments, Inc., 304 F. Supp. 2d 726 (D. Md. 2004)..................................................................................... 27 Louis Marx & Co. v. Buddy L. Corp., 453 F. Supp. 392 (S.D.N.Y. 1978) ...................................................................................... 38 Medichem, S.A. v. Rolabo, S.L., 437 F.3d 1157 (Fed. Cir. 2006) ...................................................................................... 21-22 Mobil Oil Corp. v. Amoco Chem. Corp., 779 F. Supp. 1429 (D. Del. 1991), aff'd, 980 F.2d 742 (Fed. Cir. 1992).............................. 27 Monarch Knitting Mach. Corp. v. Sulzer Morat GmbH, 139 F.3d 877 (Fed. Cir. 1998)............................................................................................. 31 Multiform Desiccants, Inc. v. Medzam, Ltd., 133 F.3d 1473 (Fed. Cir. 1998) ........................................................................................... 50 In re Omeprazole Litig., 490 F. Supp. 2d 381 (S.D.N.Y. 2007) ................................................................................. 58 Ortho-McNeil Pharms., Inc. v. Mylan Labs., Inc., 348 F. Supp. 2d 713 (N.D.W. Va. 2004), aff'd, 161 F. App'x 944 (Fed. Cir. 2005)........ 51-52 Pro-Mold & Tool Co. v. Great Lakes Plastics, Inc., 75 F.3d 1568 (Fed. Cir. 1996)............................................................................................. 31

v

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 8 of 68

Purdue Pharma L.P. v. Endo Pharms., Inc., 438 F.3d 1123 (Fed. Cir. 2006) ...........................................................................................49, 52 Ruiz v. A.B. Chance Co., 234 F.3d 654 (Fed. Cir. 2000)............................................................................................. 27 Stratoflex, Inc. v. Aeroquip Corp., 713 F.2d 1530 (Fed. Cir. 1983)+............................................................................................... 27 Takeda Chem. Indus., Ltd. v. Alphapharm Pty., Ltd., 492 F.3d 1350 (Fed. Cir. 2007) .................................................................................... Passim Takeda Chem. Indus., Ltd. v. Mylan Labs., Inc., 417 F. Supp. 2d 341 (S.D.N.Y. 2006) .................................................................................. 50-51 Yamanouchi Pharm. Co. Ltd. v. Danbury Pharmacal, Inc., 231 F.3d 1339 (Fed. Cir. 2000) ...................................................................................... 25-26 FEDERAL STATUTE 35 U.S.C. § 282 ........................................................................................................................ 21

vi

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 9 of 68

I.

PRELIMINARY STATEMENT Teva's business strategy is clear. The proton pump inhibitor ("PPI") market is large and

profitable. As a result, Teva has challenged the patentability of many branded PPIs currently on the market. By now, the nature of these challenges have become predictable: Teva routinely asserts that the discovery of the compound must have been obvious in light of the prior art, and the patent must have been procured by inequitable conduct. As the evidence at trial demonstrates, Teva hit a dry well in the case of lansoprazole. At trial and in its opening post-trial brief, Teva fails to explain why the chemical structure of lansoprazole would have been an obvious choice from among more than 13 million possible modifications to the prior-art timoprazole compound. Teva also fails to explain why one would modify omeprazole (itself a modification of timoprazole) to construct lansoprazole, which employs different substituents in different positions. In fact, the omeprazole molecule cannot be directly converted into lansoprazole. Teva's brief also ignores the testimony of Plaintiffs' expert, Dr. Lipinski, regarding the four unique hurdles to PPI design that hindered any "logical" development, the prior art which taught away from the chemical structure of lansoprazole, and the difficulties that dozens of pharmaceutical companies experienced in trying to develop a new PPI. Teva also does not rebut the secondary considerations supporting lansoprazole's novelty: Teva cannot explain away the failure of 40 or more pharmaceutical companies to develop a viable PPI, and the success of only a handful; the $34 billion commercial success of lansoprazole; or lansoprazole's unexpected advantages like faster onset of action and pain relief, fewer drug interactions, and greater solidstate stability. common sense. Teva's efforts to trivialize these objective indicia of nonobviousness defy

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 10 of 68

At core, Teva's obviousness challenge baldly contends, without record citation, that any substitution on the toxic prior-art timoprazole skeleton would work with the proton pump. Teva Br.1 at 31 ("[W]hatever the substitution pattern ... the timoprazole skeleton key still fits the Proton Pump's lock"). Dr. Lipinski's testimony showed why Teva was wrong: many variations of timoprazole unpredictably lead to compounds with potential ADME (Absorption, Distribution, Metabolism, Excretion) or toxicity problems. Lipinski 1180:7-1182:12; 1183:11184:16.2 And many variations would not be stable in the blood, convert into an active species, or fit into the proton-pump protein. Lipinski 1173:15-1174:8. Indeed, Teva's contention is belied by the efforts of the more than 40 companies that failed to develop a viable PPI, including Searle, where Teva's experts, Drs. Lenz and Dajani, worked during the relevant invention period. Teva contends that its obviousness and inequitable conduct challenges are "inextricably intertwined." Teva Br. at 1. Plaintiffs agree. Both challenges suffer from the same

shortcomings. The evidence at trial demonstrates that: (1) the inventors and Takeda management relied on the same IMAU data reported in the `098 patent to pursue lansoprazole over other novel anti-ulcer compounds; (2) the raw data in the inventors' lab notebooks confirms the accuracy of the IMAU data in the `098 patent; and (3) the "additional data" that Teva contends is evidence of inequitable conduct is either irrelevant to, or supports, the inventors' belief in the `098 compounds' superior anti-ulcer effect. In light of this record, Teva presents no evidence

that the inventors--Drs. Nohara and Maki--or anyone else who was substantively involved in the prosecution of the `098 patent, intended to mislead the PTO.
1

"Teva Br. at" refers to the corrected "Opening Post-Trial Brief of Defendants Teva Pharmaceuticals USA, Inc. and Teva Pharmaceutical Industries Ltd. Regarding the Invalidity and Unenforceability of the `098 Patent and Invalidity of the `321 Patent" submitted as D.I. 165 on December 11, 2007.
2

Citations to the trial transcript are in the format "Witness page:line-page:line." 2

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 11 of 68

The `098 patent table Teva points to also demonstrates the inventors' good faith in asserting that the compounds of the invention have a superior anti-ulcer effect "about 1.5-20 times" that of known compounds. Teva glosses over the fact that the `098 patent's results include data from two prior compounds--the "Byk Gulden compound" and omeprazole. Teva focuses on the "20-fold" difference between lansoprazole's and omeprazole's IMAU ID50 values. But the `098 patent's inventors provided the PTO with data showing only a 5.5-fold superior anti-ulcer effect between lansoprazole and the Byk Gulden compound. Dajani Cross 943:14-23. Teva's case boils down to the bare contention that the inventors should have provided the PTO each and every animal test conducted on lansoprazole. But Teva cites no law to support its contention. And Teva focuses on antisecretory testing that is distinguishable from and not comparable to anti-ulcer testing, despite the `098 patent's emphasis on its new compounds' superior anti-ulcer effects. Finally, concerning Plaintiffs' `321 formulation patent, Teva's obviousness challenge largely rehashes art that the PTO already considered, with the exception of some allegedly invalidating references Teva proffered for the first time during trial. The new references hardly salvage Teva's challenge. Teva never explains why, from among thousands of possibilities, a person of ordinary skill in the art would select magnesium carbonate as a stabilizer, and then having done so, would contact it evenly with lansoprazole in order to stabilize lansoprazole. Certainly, none of the prior art Teva cites did so, leaving Teva with nothing more than obviousness by hindsight.

3

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 12 of 68

II. A.

BACKGROUND TO ACID-RELATED GASTROINTESTINAL DISORDERS AND PROTON PUMP INHIBITORS Peptic Ulcer Disease and Gastroesophageal Reflux Disease (GERD) Drs. Maki and Nohara invented the Proton Pump Inhibitor ("PPI") lansoprazole to treat

gastric ulcers and related diseases. The stomach's "parietal cells" secrete gastric acid when the "proton pump" within these cells is triggered by one of three receptors. Fennerty 824:17-825:8. Gastric acid is extremely acidic, having a pH of 1 to 2. Id. Normally, the stomach protects itself from gastric acid through a series of cytoprotective factors. Id. However, when the stomach's cytoprotective defense mechanisms are reduced, an imbalance results that can lead to ulceration. Id. Gastric ulcers are formed in the stomach cavity itself, and duodenal ulcers are formed in the duodenum or upper small intestine. Id. Ulcers can also form in the esophagus due to gastroesophageal reflux disease or GERD. Fennerty 826:3-22. GERD results when the barrier between the esophagus and the stomach malfunctions and allows gastric acid to reflux into the esophagus, producing symptoms such as heartburn, nausea, trouble swallowing, and chest pain as well as esophageal ulcers and even esophageal cancer. Id. B. Treatment for Peptic Ulcer Disease and Gastroesophageal Reflux Disease Histamine 2 receptor antagonists or H2 blockers were the first successful, non-surgical therapy to treat ulcers. Fennerty 829:16-830:2. There are currently four H2 blockers on the market: Tagamet®, Zantac®, Pepcid®, and Axid®. Id. H2 blockers targeted the histamine

receptor, one of three known receptors that direct the proton pumps within parietal cells to produce gastric acid. Id. H2 blockers could not fully stop the production of acid because they have no effect on the other two receptors that control the proton pumps. Lenz 734:18-735:3, 735:11-24.

4

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 13 of 68

In 1989, omeprazole (trade name Prilosec®), the first commercially-successful PPI, was launched. Fennerty 830:4-831:18. Omeprazole suffered from drawbacks such as delayed onset of action, a number of drug-drug interactions, and safety concerns over carcinoid tumors that led the FDA to require a "black box" warning. Id. Moreover, although omeprazole had strong antisecretory properties, it was believed to lack cytoprotective properties. Lipinski 1163:12-16. When lansoprazole (Prevacid®) was launched in 1995, it offered an attractive alternative due to its more rapid onset of action, reduced drug-drug interactions, and lack of a black-box warning. Fennerty 847:6-12. Lansoprazole has been one of the most successful PPIs, with total sales of over $34 billion since its launch in 1995 and sales over $1 billion annually since 1998. Cole 1056:24-1058:7; PTX 182. Only three other PPIs have subsequently been introduced:

pantoprazole (Protonix), rabeprazole (Aciphex), and esomeprazole (Nexium). C. Medicinal Chemists in 1984 Knew Little About How to Develop Lansoprazole or Other Successful PPIs Drug design is very difficult with a very low success rate. Lipinski 1160:23-1161:25. In addition to the inherent biological activity of the compound, it cannot be toxic and must be safe. Id. The compound must also have adequate "ADME properties" (absorption, distribution,

metabolism, and excretion). Id.; Lenz Cross 783:3-15. Around 1984, a medicinal chemist typically would have to make an estimated 10,000 compounds before an approved compound was discovered. Lipinski 1162:3-9. Teva's expert, Dr. Lenz, conceded that more than 40 pharmaceutical companies had sophisticated medicinal chemistry groups in the 1980's and the financial desire to develop a PPI. Yet only four

companies succeeded in discovering and developing a viable PPI. Lenz Cross 777:11-18; 778:21-779:1. In fact, Dr. Lenz's former employer, Searle, unsuccessfully tried to develop a PPI using state-of-the-art computer drug design technology, but found that "changes in substituents

5

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 14 of 68

led to such unpredictable effects in a PPI's properties" that the computer technology was largely useless. Lenz Cross 779:2-781:13. III. LANSOPRAZOLE WAS A DIFFICULT AND SURPRISING DISCOVERY Dr. Lipinski testified that even persons of extraordinary skill, like himself and Dr. Lenz, would have had extreme difficulty discovering lansoprazole. Lipinski Redirect 1277:23-1278:3. Dr. Lenz conceded that it was a rare medicinal chemist who succeeded in discovering any commercially-approved product, and that he had never done so. Lenz Cross 782:16-24. The reason it was so difficult to develop a successful PPI was because, as Dr. Lenz admitted, even a single substituent change on a compound can mean the difference between the success and failure of the drug in clinical trials. Lenz Cross 785:3-15. Dr. Lenz admitted that a single change in a substituent can impact the efficacy, stability, side effects, toxicity, absorption, metabolism, distribution, and excretion of a drug, and that this was the nature of "medicinal chemistry." Id. The effort to produce PPIs by successful companies was extraordinary. For example, Byk Gulden synthesized over 650 compounds before it succeeded with its PPI, pantoprazole, after seven years of development. Lipinski 1193:1-13; PTX 33 at Lipinski 0938. In another example, esomeprazole resulted from the efforts of over 30 Astra scientists and several hundred compounds synthesized. Lipinski 1193:15-25; PTX 53 at Lipinski 0857. Both pantoprazole and esomeprazole were developed and introduced after lansoprazole. Lipinski 1193:11-13. A. Small Changes Affect Not Only Activity, but also Toxicity and ADME A compound's toxicity and ADME characteristics are highly sensitive to seemingly small changes in chemistry. For example, replacing one hydrogen in vinegar with a fluorine results in toxic fluoroacetic acid. Lipinski 1171:5-1172:7. Adding a methyl group to methanol, which is toxic, results in ethanol, which you can buy at a liquor store. Id. Finally, Tagamet, the 6

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 15 of 68

successful H2 blocker, was developed from a toxic compound that caused death during clinical trials. Id. One small substituent change transformed a toxic compound into a compound so safe that it is now sold over-the-counter. Id. B. A Team of Takeda Scientists Devoted Over Two Years and Synthesized Over 400 Compounds Before Discovering Lansoprazole Takeda Pharmaceutical Company, Limited is the largest manufacturer of pharmaceutical products in Japan. Kubo 1105:25-1106:5. Takeda sought to develop an anti-ulcer drug superior to the drugs available at the time. Kubo 1108:2-4. Takeda recognized several problems with prior-art compounds. These prior-art compounds had good antisecretory properties, but did not have any protective effect on the gastric mucosa. The compounds also were unstable, that is, subject to rapid decomposition. Lipinski 1163:14-24; PTX 1 at Col. 1, ll. 13-19. Takeda thus tried what more than 40 other pharmaceutical companies also tried--to develop an anti-ulcer drug that both inhibited gastric acid secretion and enhanced protection of the gastric mucosa. Lipinski 1163:22-24; PTX 1 at Col. 1, ll. 23-26; PTX 44 at Lipinski 0799. In 1982, Mr. Keiji Kubo began work on Takeda's anti-ulcer project as part of Dr. Akira Nohara's synthesis group. Kubo 1107:7-15. Dr. Nohara's group synthesized hundreds of different

chemical compounds in its effort to develop a novel anti-ulcer agent. Another Takeda group, led by Dr. Yoshitaka Maki, evaluated the biological properties of compounds synthesized by Dr. Nohara's group. Kubo 1108:5-19. These groups pursued many synthesis strategies to develop an anti-ulcer drug and failed many times before ultimately succeeding with lansoprazole. 1. Takeda Scientists Synthesized a Lead Compound: Timoprazole

The inventors did not have an initial synthesis strategy to develop an anti-ulcer compound. Through trial and error, the inventors developed compounds and then tested them for

7

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 16 of 68

anti-ulcer activity. Kubo 1113:18-1114:5. As a result, the inventors had no lead compounds when the anti-ulcer project began. Instead, the inventors conducted random screening on

compounds already synthesized for other projects to determine if those compounds had any antiulcer activity. Kubo 1109:7-13. Ultimately, the inventors settled on a compound identified internally as AA-1569 as their lead compound for its anti-ulcer project. AA-1569 was also known as timoprazole (PTX 60 at Lipinski 001476):
C C C C C C N H N C C S O CH2 C N C C C

Takeda scientists originally synthesized timoprazole to determine its potential as an allergy medicine in January 1982. Kubo 1109:21-24, 1113:3-17; PTX 639 at 446543. Through testing, the inventors discovered that timoprazole showed promise as an anti-ulcer agent (identified as AG-879). What the inventors did not know at the time was that timoprazole was already known in the art, Kubo 1109:17-1110:5, and, as Dr. Lenz admitted, that timoprazole was toxic. Lenz Cross 787:22-799:2. The inventors also learned of omeprazole at some point during their anti-ulcer compound development project. Omeprazole contained the same timoprazole skeleton that the inventors had been working with:

8

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 17 of 68

Omeprazole
Methoxy Methyl
O

Methoxy

O N

N N O

Methyl

Benzimidazole

Pyridine

Takeda scientists recognized that they needed to find a compound superior to omeprazole that did not have the toxicity of the timoprazole skeleton. Omeprazole therefore became one of the benchmarks in their efforts, and work on the timoprazole skeleton proceeded. 1121:12-21; PTX 61 at Takeda 436712. 2. The Inventors' Efforts to Modify Timoprazole Kubo

The five PPIs commercially-available today (omeprazole, lansoprazole, pantoprazole, rabeprazole, and esomeprazole) all share the timoprazole skeleton. Lipinski 1166:2-16. Due to timoprazole's toxicity, the need to modify it was critical. But how to do so? By using 20 common substituents selected from the `431 omeprazole compound patent, both Dr. Lipinski and Dr. Lenz agreed that there are over 13 million possible compounds that all shared the timoprazole skeleton, none of which is lansoprazole. Lipinski 1169:1-23; DTX 5; Lenz Cross 789:15-790:7. Indeed, one of the substituents that Takeda scientists ultimately discovered was useful, trifluoroethoxy, was not disclosed as a substituent in the `431 patent. Lenz Cross 792:20793:8. After selecting timoprazole as the lead compound, the inventors' strategy was to experiment by changing one or more of the three components of the timoprazole skeleton: the benzimidazole portion, the methylsulfinyl portion, and the pyridine portion. Kubo 1109:171110:5, 1115:21-25; PTX 45 at Takeda 436696.

9

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 18 of 68

This approach was unsuccessful.

The inventors eventually determined that the

unmodified timoprazole skeleton was useful for the compound to have anti-ulcer activity, and placed less emphasis on trying to modify the timoprazole skeleton itself. Kubo 1119:16-21; PTX 45 at Lipinski 001263.3 Takeda scientists began experimenting with the addition of various substituents to the timoprazole skeleton. PTX 61 at Takeda 436712. But they had no indication of which particular position on the timoprazole skeleton to start with, because there was no logical place to start. Kubo 1122:11-15. For example, they tried a number of substitutions on the benzimidazole ring, including several fluorinated substituents.4 Kubo 1124:20-1125:3; PTX 60 at Lipinski 01582, 01559. But these scientists discovered that substituents on the benzimidazole ring did not increase anti-ulcer activity. Kubo 1124:17-21, 1125:6-18; PTX 61 at 436713. In fact,

lansoprazole, unlike omeprazole, has no substituents on the benzimidazole ring.

3

Yet, even after lansoprazole was synthesized, the anti-ulcer team at Takeda continued to create and test modifications of the timoprazole skeleton. Kubo 1118:8-13. Although many more examples exist, five examples of this strategy are: AG-1758, AG-1779, AG-1780, AG1781 & AG-1782. PTX 60 at LIPINSKI 001586, 001588-89.
4

Although many more examples exist, five examples of this strategy are: AG-1545, AG-1555, AG-1565, AG-1567 & AG-1575. PTX 60 at LIPINSKI 001559-60, 001562-63. 10

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 19 of 68

The inventors also tried numerous substituents at the 4-position of the pyridine ring. 5 Kubo 1124:4-10; PTX 60 at Lipinski 01564, 01580. In November 1983, 18 months after they first synthesized timoprazole, the inventors reported that a compound with a long alkoxy chain on the pyridine ring (AG-1511) showed superior anti-ulcer properties. 1124:15-16; PTX 70 at Takeda 040875. The inventors thus began to focus on the 4-position in the pyridine ring. They Kubo 1123:14-21,

experimented by adding trifluoroethoxy substituents to the timoprazole skeleton, and around October 1983, they synthesized AG-1598. Kubo 1126:8-1127:9; PTX 641 at Takeda 446784. Due to AG-1598's activity and stability, the inventors endeavored to enhance the activity of AG1598 and introduced various other substituents onto it before coming up with lansoprazole. Kubo 1126:15-19; Lenz Rebuttal 1437:7-9; 1439:9-11. Finally, in May 1984, 28 months after it first synthesized timoprazole, Dr. Kubo (then Mr. Kubo), inventor Dr. Nohara's assistant, synthesized lansoprazole. Kubo 1127:15-23; PTX 648 at Takeda 039840:

Lansoprazole
Methyl
O

Trifluoroethoxy
F F F N

N N O

Benzimidazole

Pyridine

It took the inventors 28 months to proceed from the timoprazole lead compound to lansoprazole. Kubo 1128:14-17. During this 28-month period, the inventors and their assistants

5

Although many more examples exist, five examples of this strategy are: AG-1511, AG-1586, AG-1587, AG-1598 & AG-1599. PTX 60 at LIPINSKI 001555, 001564, 001566. 11

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 20 of 68

synthesized more than 400 compounds in their search for a novel anti-ulcer agent. 1128:23-1129:2; PTX 60 passim.

Kubo

In August 1984, well before the commercial release of omeprazole, the inventors filed Japanese Patent Application No. 59-171069 for lansoprazole and related compounds. On July 29, 1985, the inventors filed United States Patent Application 760,568, which matured into U.S. Patent No. 4,628,098 (the `098 patent). PTX 1. The Patent Office allowed the `098 patent during prosecution on the first office action. Lipinski 1165:17-23; DTX 2 passim. Claim 10 of the `098 patent claims 2-[3-methyl-4-(2,2,2,-trifluoroethoxy)-pyrid-2-

yl]methylsulfinylbenzimidazole, or lansoprazole. Lipinski 1154:21-155:2; PTX 1 at Col. 10. C. Four Unique Hurdles Make PPI Development Unpredictable PPI chemistry is unusually complex. Even today, the chemistry and mechanism of action of the proton pump inhibitor (PPI) class of drugs is considered one of the most novel and unusual in the history of the pharmaceutical industry, despite the existence of PPIs for over 30 years. Lipinski 1172:16-1173:8; PTX 18 at 365. Unlike other drugs, in PPI development, there are four additional and unique hurdles one must overcome for the drug to be successful. Lipinski 1173:15-1174:8. 1. Hurdle 1: The PPI must accumulate solely in the parietal cell

PPIs must accumulate solely in the parietal cell. If the drug accumulates in other tissues in the body, toxicity may result. Lipinski 1174:12-1175:15. It was understood in 1984 that accumulation in other tissues was a hurdle to PPI development. Teva makes much of the fact that DTX 13, a 1984 editorial, disclosed a preference to restrict pKa to less than 4.5, and that a skilled scientist would prefer a range between 3 and 4.5. Lenz Cross 807:17-22, 808:1-5, 807:49, 807:13-16.

12

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 21 of 68

But knowing about accumulation alone would not provide a reasonable expectation of success, or show the way to choose the correct modifications, to develop a successful PPI. In fact, the omeprazole `431 patent did not even mention pKa or describe its importance. Lenz Cross 806:20-24. The range of pKas between 3 and 4.5 is enormous (it is a log scale), and only slightly reduces the millions of possibilities for how to modify the timoprazole skeleton. Lipinski 1175:7-11. 2. Hurdle 2: The PPI must be stable in the blood (i.e., at neutral pH)

Teva ignores the second hurdle: after ingestion, the PPI must be absorbed from the gastrointestinal tract and travel through the blood to reach the parietal cell without breaking down. In other words, the PPI must have neutral pH stability. Lipinski 1175:12-1176:22. In 1984, people knew there was a problem concerning stability of PPIs at neutral pH, but they had no idea what the solution was, as demonstrated by PTX 30, a 1992 paper showing that scientists were still grappling with the issue. Lipinski 1176:8-11. There was no logical way in 1984 to identify particular substituents to address this hurdle. 3. Hurdle 3: The PPI must convert to an active species in the body

All PPIs are prodrugs. PPIs like lansoprazole convert to an active species (the "PPI active species") in the body. This PPI active species then inhibits the proton pump. Not all timoprazole-like compounds will successfully undergo this conversion. Lipinski 1179:12-15. In 1984, the chemistry for this conversion process was unknown, as demonstrated by PTX 31 and PTX 32, two papers from pharmaceutical companies, Upjohn and Hoechst, in 1985 that had the chemistry all wrong. Lipinski 1178:23-1179:10. In fact, the chemistry of this conversion process was worked out for the first time in PTX 17, a 2004 paper (20 years after lansoprazole was synthesized), and even this paper has not solved all of the mystery behind this chemistry. Lipinski 1179:16-1180:6. 13

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 22 of 68

4.

Hurdle 4: The PPI active species must fit the proton pump Lipinski 1180:10-

The proton pump is a very complex, three-dimensional protein.

1182:12. For inhibition of gastric acid to occur, the PPI active species must fit into the protein, much like a key fits into a lock. Id. The substituents on the timoprazole skeleton determine whether there will be a good fit. Id. For example, the three-dimensional structures of

lansoprazole and omeprazole are very different, and thus the fit of their corresponding PPI active species into the proton pump will differ. Lipinski 1166:24-1167:24. The substituents on the timoprazole skeleton determine how well the PPI active species fits into the proton pump, and thus determine in part the activity. Lipinski 1181:18-22. There was no way to know in advance what substituents would help or hurt this fit. Lipinski 1181:23-1182:12. Even today, medicinal chemists do not fully understand the structure of the proton pump. Id. Teva's expert, Dr. Lenz, did not testify about any prior art that addressed three of these unique hurdles, Lipinski 1182:13-25, 1184:17-21, even on rebuttal. D. Prior Art Would Have Led Medicinal Chemists Away from Developing Lansoprazole As Dr. Lipinski explained, and Teva's Dr. Lenz ignored, the principal prior art references Teva relies on actually teach away from the structure of lansoprazole. 1. The `431 Astra Patent

U.S. Patent No. 4,255,431 (DTX 5) claimed omeprazole and was considered by the PTO during the `098 patent prosecution. Lipinski 1164:10-15, 1195:11-15; DTX 5. The `431 patent taught a preferred substitution on the pyridine ring of omeprazole: 3methyl, 4-methoxy, and 5-methyl. Lipinski 1195:19-1196:12. Examples 11, 17, 20, 21, 23, 24, 25, and 28 in Table 2 of the `431 patent all share this substitution pattern. Id.; DTX 5 at TVL

14

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 23 of 68

011025-26; Teva Br. at 20. Lansoprazole shares only the 3-methyl substitution with the `431 patent's preferred substitution pattern. Lipinski 1196:14-1197:1.
Comparison of Omeprazole and Lansoprazole Structures
Omeprazole
Methoxy Methyl
O

Methoxy

O N

N N O

Methyl

Benzimidazole

Pyridine Trifluoroethoxy Methyl
O F N F F

Lansoprazole

N N O

Benzimidazole

Pyridine Lipinski087

In fact, there is no chemical means to convert omeprazole to lansoprazole. Lipinski 1187:3-9. The `431 patent also taught that a shorter (1 carbon) substituent is preferred to a longer (2 carbon) substituent at the 4-pyridine position. Lipinski 1197:3-1198:15; DTX 5 at TVL 011026. Lansoprazole has a longer two carbon substituent at the 4-pyridine--trifluoroethoxy--precisely the kind of group the `431 patent taught away from. Lipinski 1198:8-15. Indeed, the `431 patent stressed 4-methoxy's importance, Lipinski 1196:20-22, while it did not even mention trifluoroethoxy. Lenz Cross 793:4-8. The `431 patent also taught away from removing the 5-methyl on the pyridine ring, since its activity table showed that when 5-methyl was removed, the compounds were less active. Lipinski 1198:17-1199:15; DTX 5 at TVL 011026. Lansoprazole does not have a 5-methyl on the pyridine ring. Id. Dr. Lenz never explained why one would move from 5-methyl on the pyridine ring (as in omeprazole) to a compound with no substituent at the 5-position (as in lansoprazole). Lipinski 1186:12-19.

15

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 24 of 68

Finally, the `431 patent taught away from having an unsubstituted benzimidazole ring. Of the 29 examples in Table 2 in the `431 patent, 28 of them contain substituted benzimidazole rings. Lipinski 1199:17-1200:7; DTX 5 at TVL 011025-26. In particular, the 5-methoxy

substitution in the benzimidazole ring was later found to make the omeprazole compounds much more stable to conversion at neutral pH (PPIs' unique hurdle 2). Lenz Cross 790:23-791:20; PTX 44 at Lipinski 0789. Lansoprazole has an unsubstituted benzimidazole ring, contrary to the teachings of the `431 patent. Lipinski 1168:2-20; 1196:14-19. Again, Dr. Lenz provided no explanation as to why one would go from using a substituted benzimidazole compound to an unsubstituted benzimidazole compound like lansoprazole. Lipinski 1186:2-11. In its post-trial brief, Teva notes that one compound listed in Table 2 of the `431 patent had an unsubstituted benzimidazole ring (example 25). Teva Br. at 44-45. Dr. Lenz never addressed this point at trial. But that is hardly surprising. The unsubstituted example 25 is less potent than the 28 other substituted examples. Thus, the `431 patent taught one of ordinary skill in the art to substitute the benzimidazole ring. Teva also contends that "virtually all" compounds disclosed in the `431 patent would have led to a successful PPI. Teva Br. at 35. But only the omeprazole compound described in the `431 patent was successful. One of skill in the art would have no idea whether the other compounds in the `431 patent would have been successful PPIs. Lipinski 1184:9-12. In fact, in 1984, it was uncertain that even omeprazole would be successful. Lenz Rebuttal Cross 1441:1-7. As Dr. Lenz admitted, adding or changing substituents in the omeprazole structure might produce unsafe or toxic compounds. Lenz Cross 793:9-15. 2. The `409 Byk Gulden Patent

U.S. Patent No. 4,472,409 (DTX 7) was considered by the PTO during prosecution of the `098 patent. Lipinski 1164:10-15. Like the omeprazole `431 patent, DTX 5, the `409 patent 16

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 25 of 68

taught solely compounds with a methoxy group at the pyridine ring's 4-position, so that a person of ordinary skill in the art would be taught away from replacing this methoxy group as the `098 inventors did. Lenz Cross 794:11-18; Lipinski 1200:18-1201:1 & 1201:24-1202:3. The `409 patent does not even mention trifluoroethoxy as a potential substituent. 1203:1; Lenz Rebuttal Cross 1445:1-3. Lipinski 1202:22-

Byk Gulden '409
F F N F N

Methoxy

O

N O

Benzimidazole

Pyridine

Furthermore, like the omeprazole `431 patent, the `409 patent's compounds are all substituted on the benzimidazole ring, so a person of ordinary skill in the art would be taught to maintain some substitution on the benzimidazole ring--and would be taught away from leaving it unsubstituted as in lansoprazole. Lipinski 1201:2-6. 3. The `518 Byk Gulden Patent

U.S. Patent No. 4,555,518's title--"Fluoroalkoxy Substituted Benzimidazoles Useful as Gastric Acid Secretion Inhibitors"--makes clear what this patent taught. DTX 9 (emphasis added). The `518 patent's compounds all have a fluoroalkoxy on the benzimidazole ring, not the pyridine ring. Lipinski 1203:13-23. A medicinal chemist would view the benzimidazole ring and pyridine ring in isolation, and would not switch substituents between these different ring systems. Lipinski 1203:2-12. Therefore, the `518 patent's fluoroalkoxy substituents on the benzimidazole ring would not suggest putting the same substituents on the pyridine ring. Id.

17

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 26 of 68

Like the `409 and `431 patents, the `518 patent taught a preferred methoxy substitution at the 4-pyridine position. Lipinski 1204:3-18; DTX 9 at Col. 24, ll. 15-59; 794:19-795:4. Thus, like the other patents Teva relied upon, the `518 patent taught away from replacing the 4methoxy group. In short, the `518 patent did not teach the use of trifluoroethoxy in the pyridine ring. Lipinski 1203:20-23, 798:2-8. Lansoprazole does not have a methoxy group at the 4-pyridine position, as the `431, `409, and `518 patents required. And even Dr. Lenz admitted that he would not have shifted a trifluoromethoxy, preferred in the `518 patent, from the benzimidazole ring to the 4-position of the pyridine ring, which shows how these rings differ. Lenz Cross 815:4-17. Finally, Teva does not dispute that no commercially viable products resulted from the `409 or `518 patents, because the compounds disclosed in those patents suffered from neutral-pH instability or toxicity problems. Lenz Cross 795:6-15; 1165:1-10; 1201:7-23. E. Medicinal Chemists Would Not Have Selected Trifluoroethoxy as a Substituent in 1984 Lansoprazole was synthesized with a trifluoroethoxy at the 4-position of the pyridine ring. This was a novel step. Teva argues that the use of "fluorinated substituents," like

trifluoroethoxy, would have been obvious because of "fluorine's well-known properties." Teva Br. at 39-40. Teva grossly overreaches in several respects. See, e.g., Lipinski 1171:12-14. "Fluoroethoxy" refers to an entire family of substituents having hundreds of members. Lipinski 1187:20-25; 1188:7-13. "Trifluoroethoxy" refers to a family of substituents having three

members. Lipinski 1188:14-21. Lansoprazole has a particular trifluoroethoxy, called the "2,2,2trifluoroethoxy." Lipinski 1188:22-1189:6. If 2,2,2-trifluoroethoxy were replaced by say 1,2,2trifluoroethoxy, the result would be completely unpredictable, but most likely would not lead to a successful compound. Id.

18

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 27 of 68

Moreover, as Dr. Lipinski testified and his own investigation confirmed, far from being a logical or obvious choice, trifluoroethoxy was in fact an unusual "fluorinated substituent" to choose in 1984. As of August 16, 1984 (the date the patent application was filed in Japan), there were only two chemical compounds known with a trifluoroethoxy at the 4-position of a pyridine ring out of around 30 million compounds searched under Dr. Lipinski's direction. Lipinski 1189:12-1190:10, 801:5-802:13; PTX 37 at Lipinski 001325. One of the compounds was an herbicide and the other was an anti-malarial compound. In fact, only four compounds in clinical trials as of August 1984 contained a trifluoroethoxy at any position of the molecule. Lipinski 1190:11-1192:2, 801:5-802:11; PTX 38 at Lipinski 001335. None of those four identified compounds was a gastrointestinal drug. Dr. Lipinski also pointed out that a 1991 reference compiling information on 530 common substituents did not even include trifluoroethoxy in its compilation. Lipinski 1191:10-1192:3; 800:13-21; DTX 101 at TVL 63322-63329. The use of trifluoroethoxy could hardly be considered an obvious substituent choice in 1984. In fact, lansoprazole is the only one of the five available PPIs that has a fluorine at any position on the pyridine ring. Lenz Rebuttal Cross 1446:11-13; Lipinski 1191:25-1192:3. F. Teva's Reliance on DTX-703 is Baseless Teva's brief relies heavily on DTX 703 for its obviousness positions. See, e.g., Teva Br. at 2. DTX 703 contains the conclusory statement that the "structure [of lansoprazole] is too similar to Omeprazole." But there is no evidence concerning who authored DTX 703, nor the author's educational background or position, and no evidence concerning even its date or circumstances of creation. There is no evidence concerning the relationship between DTX 703's author and the inventors, or anyone else substantively involved in the `098 patent's prosecution. No one knows if the inventors or anyone else even read DTX 703. Dr. Kubo, the first person to synthesize lansoprazole, testified that he was unfamiliar with DTX 703. Kubo Cross 1149:2319

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 28 of 68

1150:9. Indeed, DTX 703 on its face appears to be a draft or uncompleted document, as a number of entries on it were not completed. See Teva Br. at 48 (title, date, rater, etc., blank). Moreover, Dr. Kubo testified that he had never heard anyone at Takeda express the view that the inventors could not get a patent on lansoprazole because of omeprazole, contrary to Teva's interpretation of DTX 703. Kubo Redirect 1154:1-4. Teva has no evidence to the contrary. IV. SUBSTANTIAL DIFFERENCES, SCOPE AND CONTENT OF THE PRIOR ART, LEVEL OF ORDINARY SKILL, AND OBJECTIVE INDICIA SHOW THAT LANSOPRAZOLE WAS NONOBVIOUS Teva contends that a medicinal chemist in 1984 would have found lansoprazole an obvious and "logical" progression from the prior art.6 Teva Br. at 20. But, as described in detail above, the testimony and documents in evidence refute this argument. First, Teva fails to demonstrate how one of skill in the art in 1984 would have a reasonable expectation of success by following the prior art teachings to construct lansoprazole. Second, Teva tries to construct lansoprazole from omeprazole without any explanation of why one of skill in the art would choose to make the necessary modifications to omeprazole to construct lansoprazole. Teva's conclusory construction is a paragon of hindsight, which the Federal Circuit and Supreme Court have expressly warned against in an obviousness analysis. Third, Teva completely ignores the prior art's teaching away from Teva's construction of lansoprazole. Finally, Teva ignores all objective indicia evidence that demonstrated lansoprazole's non-obviousness.

6

Teva's confidence in its obviousness position is belied by the fact that their initially filed ANDA in August 2004 included a Paragraph III certification indicating they would not contest the validity of the `098 patent. 20

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 29 of 68

Under 35 U.S.C. § 282, each claim of the `098 patent is presumed valid. "Any facts supporting a holding of invalidity must be proved by clear and convincing evidence." Budde v. Harley-Davidson, Inc., 250 F.3d 1369, 1376 (Fed. Cir. 2001). Teva suggests, without any legal support, that an invention requires a "Eureka moment" or a "light-bulb moment" to be non-obvious. Teva Br. at 22. Teva is wrong. "It is immaterial whether the invention resulted from long toil and experimentation or from a flash of genius." CFMT, Inc. v. Yieldup Int'l Corp., 349 F.3d 1333, 1340 (Fed. Cir. 2003) (citations omitted). Even "systematic variation" does not make compounds obvious. Bayer AG v. Dr. Reddy's Labs., Ltd., No. 04-179-SLR, 2007 WL 3120794, at *8 & n.30 (D. Del. Oct. 25, 2007). Although obviousness is a question of law, four factual underpinnings control an obviousness inquiry. These underpinnings are "[a] the scope and content of the prior art are ... determined; [b] differences between the prior art and the claims at issue are ... ascertained; [c] the level of ordinary skill in the pertinent art [is] resolved" and [d] evidence of objective indicia of nonobviousness is evaluated. KSR Int'l Co. v. Teleflex Inc., 127 S. Ct. 1727, 1729-30 (2007). A. Teva Fails to Describe a Reasonable Expectation of Success In evaluating these four factual underpinnings, a court may find obviousness only if it finds a reasonable expectation of success in arriving at the invention by combining prior art. Takeda Chem. Indus., Ltd. v. Alphapharm Pty., Ltd., 492 F.3d 1350, 1360-62 (Fed. Cir. 2007); Eli Lilly and Co. v. Zenith Goldline Pharm., Inc., 471 F.3d 1369, 1377 (Fed. Cir. 2006). A "reasonable expectation of success" requires more than "vary[ing] all parameters or try[ing] each of numerous possible choices until one possibly arrived at a successful result, where the prior art gave either no indication of which parameters were critical or no direction as to which of many possible choices is likely to be successful." Medichem, S.A. v. Rolabo, S.L., 437 F.3d 1157, 1165 (Fed. Cir. 2006) (internal citations omitted). Similarly, prior art fails to provide the requisite 21

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 30 of 68

"reasonable expectation" of success where it teaches merely to pursue a "general approach that seemed to be a promising field of experimentation." Id. Teva contends that there was an "obvious path to lansoprazole" through a "logical, methodical exercise" from the prior-art compounds. Teva Br. at 21, 40. But if such a road map existed, Teva does not explain why more than 40 companies attempted but failed to produce a viable PPI product. Nor can Teva explain why Takeda scientists synthesized more than 400 compounds before arriving at lansoprazole. As Dr. Lipinski explained at trial, there was no "logical" progression from the prior art to lansoprazole given the unpredictability of substituents added onto the timoprazole skeleton and the complexity of PPI chemistry, which prevented accurately predicting how substituent changes would affect the reaction of a compound in each of the four unique PPI hurdles. Lipinski 1159:23-1160:7, 1208:15-24. Teva's expert, Dr. Lenz, agreed with Dr. Lipinski that changing substituents would have unpredictable results in terms of a compound's ADME characteristics and toxicity. Lenz Cross 785:3-15. In fact, Dr. Lenz referred to this unpredictability as being at the heart of "medicinal chemistry," the relevant art in this case. Id. Teva does not, and cannot, dispute that changes to substituents in timoprazole resulted in unpredictable results in terms of the compound's ADME and toxicity characteristics. There was no reasonable expectation of success that the `098 inventors' choice of substituents in their particular positions on the timoprazole skeleton would have worked. Teva never addresses why one of ordinary skill in the art would find a reasonable expectation of success by using a trifluoroethoxy substituent, as in lansoprazole, particularly at the 4-pyridine position. Teva does not refute that trifluoroethoxy was a highly unusual choice of substituent, and that trifluoroethoxy at the 4-pyridine position was extremely rare for all

22

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 31 of 68

chemicals. See Lipinski 1189:12-1190:10, 801:5-802:11; PTX 37 at Lipinski 001325. Nor did Teva respond at trial to Dr. Lipinski's opinion that at least three of the hurdles in PPI chemistry prevented one of skill in the art in 1984 from accurately predicting how a compound's substituents would work in the body. Thus, a review of the scope and content of the prior art shows that lansoprazole's substituents were unusual, and that the expectation of success, especially in light of the four unique hurdles, was low. B. Teva's Purported "Obvious Path to Lansoprazole" Uses Hindsight and Ignores the Prior Art's Teaching Away "A critical step in analyzing the patentability of claims pursuant to section 103(a) is casting the mind back to the time of invention, to consider the thinking of one of ordinary skill in the art, guided only by the prior art references and the then-accepted wisdom in the field." Ecolochem, Inc. v. S. Cal Edison Co., 337 F.3d 1361, 1371 (Fed. Cir. 2000). The Federal Circuit has warned against using "the inventors' disclosure as a blueprint for piecing together the prior art to defeat patentability -- the essence of hindsight." Id. at 1371-72. As the Federal Circuit and Supreme Court have made clear, to establish a prima facie obviousness case, it remains necessary, even after KSR, to describe what would have led an ordinary chemist to modify a known compound in a particular manner. Takeda v. Alphapharm, 492 F.3d at 1357; KSR, 127 S. Ct. at 1741 (requiring courts "to determine whether there was an apparent reason to combine the known elements in the fashion claimed by the patent at issue."). To justify its theory that one of ordinary skill in the art would have a reason to modify omeprazole to construct lansoprazole, Teva uses hindsight to work backward from lansoprazole to omeprazole. Teva first argues that one of skill in 1984 would somehow know to remove the substituent from the benzimidazole ring in omeprazole. But 28 of the 29 compounds disclosed in

23

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 32 of 68

the omeprazole `431 patent had a substituted benzimidazole ring, and the 29th (the sole unsubstituted compound) had less activity than the comparable substituted counterpart. Lipinski 1186:6-11. Teva's arguments regarding substituents in the pyridine ring fare no better. Teva argues that it would have been "logical" to replace the methoxy on the 4-pyridine position on omeprazole with a trifluoroethoxy, and to remove one methyl from the 5-pyridine position from omeprazole to construct lansoprazole. But none of Teva's references taught the use of a

trifluoroethoxy substituent at any position on the pyridine ring of timoprazole, nor was there any evidence that this substitution was within the skill of one of ordinary skill in the art. It is undisputed that changing the substituent at the pyridine 4-position has a profound effect on the pyridine ring as a whole. Lenz 753:24-754:1. Yet all of Teva's references taught that a methoxy at the pyridine 4 position was essential, and thus taught away from replacing this 4-methoxy with any other group. Lipinski 1200:18-1201:1, 1204:3-18. In fact, the `431 patent taught that a shorter 1-carbon substituent like methoxy is preferred to a longer 2-carbon substituent like trifluoroethoxy at the pyridine 4-position, thus showing how far this reference went in teaching away from lansoprazole. Lipinski 1197:3-1198:15. Because each substituent is critical, Lenz Cross 781:11-13 & 785:3-15, Lipinski 1169:241170:3, 1170:23-25 & 1171:5-1172:7, and there are at least 3 substituents described above that differ, the differences between lansoprazole and the prior art are substantial. Yet in regard to each required modification, the `431, `409, and `518 patents teach away from Teva's proposed modifications. Teva offers no prior art that taught towards lansoprazole's chemical structure.

24

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 33 of 68

C.

The Federal Circuit and this Court Have Rejected Similar Obviousness Theories Based on Similar Facts Teva's brief erroneously suggests that "structural similarity" is sufficient to establish a

prima facie case of obviousness. Teva Br. at 23. As described above, there were significant differences in the structure of lansoprazole and the prior art, and no teaching in that art to make the structural changes made by the `098 inventors to construct lansoprazole. Moreover, the Federal Circuit has clarified that while "structural relationships may provide the requisite motivation or suggestion to modify known compounds to obtain new compounds ... in order to find a prima facie case of unpatentability in such instances, a showing that the prior art would have suggested making the specific molecular modifications necessary to achieve the claimed invention was also required." Takeda v. Alphapharm, 492 F.3d at 1355-56 (internal citations omitted) (emphasis added). "For a chemical compound, a prima facie case of obviousness requires `structural similarity between claimed and prior art subject matter ... where the prior art gives reason or motivation to make the claimed compositions.'" Yamanouchi Pharm. Co. Ltd. v. Danbury Pharmacal, Inc., 231 F.3d 1339, 1343 (Fed. Cir. 2000) (emphasis added). As demonstrated above, Teva does not demonstrate how a person of ordinary skill in the art would have been motivated to modify specifically the structure of omeprazole to construct the structure of lansoprazole. The holding in Takeda v. Alphapharm is instructive. The difference between the prior art and the patented compound was a 6-pyridine methyl group (CH3) versus a 5-pyridine ethyl group (CH2CH3). Takeda v. Alphapharm, 492 F.3d at 1354. Alphapharm suggested that one of skill in the art would routinely convert the methyl to an ethyl and "ring walk" the ethyl group from the 6-pyridine position to the 5-pyridine position to construct the patented compound. Id. at 1360-

25

Case 1:06-cv-00033-SLR

Document 173-2

Filed 01/10/2008

Page 34 of 68

61. The court disagreed and noted that even Alphapharm's expert conceded that the "biological activities of various substituents were unpredictable." Id. at 1361. See also Yamanouchi, 231 F.3d at 1341 (structurally similar prior-art compound does not make new compound obvious). Alphapharm's expert's "ring-walking" theory is analogous to Dr. Lenz's unsupported move of a 2,2,2-trifluoroethoxy, only one of many possible substituents in the Byk Gulden `518 patent's choices on the benzimidazole ring, from the 5-position of the benzimidazole ring to the 4-pyridine ring position. Even worse, to construct lansoprazole, Dr. Lenz then further removed and replaced other substituents on the pyridine ring. This Court's recent decision in Bayer v. Dr. Reddy's Labs. is similarly instructive. The Court found that Reddy did not demonstrate that the prior art taught modification of the lead compound with the necessary substituents to create the claimed compound, or that there was a reasonable expectation of success to create this compound. Bayer, 2007 WL 3120794, at *9. Yet in this case, Teva's prior art, unlike the prior art in Yamanouchi, Bayer, or Takeda, actually taught away