Janssen Prods., L.P. v. Lupin Ltd.

Full title: JANSSEN PRODUCTS, L.P., et al., Plaintiffs, v. LUPIN LIMITED, et al.…

Court: United States District Court, D. New Jersey.

Date published: Aug 14, 2014

109 F. Supp. 3d 650 (D.N.J. 2014)

Opinion

No. 2:10–CV–05954 WHW.

08-14-2014

JANSSEN PRODUCTS, L.P., et al., Plaintiffs, v. LUPIN LIMITED, et al., Defendants.

Zhiqiang Liu, Patterson Belknap Webb & Tyler, New York, NY, Carissa L. Rodrigue, Elina Slavin, John Edmund Flaherty, Jonathan M.H. Short, Mark H. Anania, Ravin R. Patel, McCarter & English LLP, Newark, NJ, for Plaintiffs. G.D. Searle, Newark, NJ, pro se. James E. Cecchi, Melissa E. Flax, Carella Byrne Cecchi Olstein Brody & Agnello, P.C., Roseland, NJ, Mayra Velez Tarantino, Michael E. Patunas, Lite Depalma Greenberg, LLC, William C. Baton, Saul Ewing LLP, Newark, NJ, for Defendants.

WALLS, Senior District Judge.

Zhiqiang Liu, Patterson Belknap Webb & Tyler, New York, NY, Carissa L. Rodrigue, Elina Slavin, John Edmund Flaherty, Jonathan M.H. Short, Mark H. Anania, Ravin R. Patel, McCarter & English LLP, Newark, NJ, for Plaintiffs.

G.D. Searle, Newark, NJ, pro se.

James E. Cecchi, Melissa E. Flax, Carella Byrne Cecchi Olstein Brody & Agnello, P.C., Roseland, NJ, Mayra Velez Tarantino, Michael E. Patunas, Lite Depalma Greenberg, LLC, William C. Baton, Saul Ewing LLP, Newark, NJ, for Defendants.

OPINION

WALLS, Senior District Judge.

By complaints, Plaintiff patent-holders and drug-makers have sued Defendant generic drug manufacturers for alleged infringements of patents on a chemical compound used in the manufacture of Plaintiff's product, Prezista. The patents, for this opinion, are U.S. Patent No. 7,700,645 B2 (the "'645 Patent"), U.S. Patent No. 7, 126,015 B2 (the "'015 Patent"), and U.S. Patent No. 7,772,411 B2 (the "'411 Patent").

Prezista is an ethanolate form of the chemical compound named darunavir. The '015 Patent claims processes for manufacturing a chemical structure or moiety, bis-THF, that is part of the darunavir molecule. The ' 411 Patent is directed to a process for manufacturing darunavir in Prezista and Defendant Mylan's generic version. The '645 Patent claims the ethanolate form that Plaintiff Janssen developed and sells as Prezista.

Earlier summary judgment motion practice by the parties caused the Court to determine that:

1) Janssen's motion for summary judgment of infringement of the '411 Patent against Mylan was granted.

2) Janssen's motion for summary judgment on the validity of the '645 Patent was denied.

3) Janssen's motion for summary judgment of infringement of the '015 and '408 Patents was granted in part and denied in part: the Court found that Lupin infringed certain claims of the '015 Patent, including asserted claim 1.

4) Teva's motion for summary judgment of non-infringement of the '645 Patent was denied.

5) Mylan's motion for summary judgment of non-infringement of the '645 Patent was denied.

6) Lupin's and Teva's motion for summary judgment of non-infringement of the '015 and '408 Patents was denied.

Before trial, Teva and Janssen resolved their dispute, and the '408 Patent is no longer at issue.A bench trial has been held on the remaining matters: the validity of the patents—as to Lupin, claim 4 of the '645 Patent and claim 1 of the ' 015 Patent, and as to Mylan, claim 13 of the '411 Patent.

This Opinion will be based and concentrated on what the Court has found to be material and supportive facts necessary to answer the primary question: Did either Defendant prove invalidity of a patent-in-suit by clear and convincing evidence?

To find the answer(s), the Court has reviewed the relevant evidence adduced at trial, aided by recollection, notes, trial transcripts, and the parties' proposed findings of fact. The Court has evaluated the credibility of all witnesses, lay and expert, testing not only what the person said, but how it was said and what was not said, against these criteria: Was this more likely so or not? Did what was said make sense in the totality of the circumstances? Was this, after all, clearly convincing to the fact finder?

To its chagrin, the Court has determined that expert witnesses, though not all, called by the defense were more interested in obfuscation than in helping the Court as a fact finder "seek the truth." Too often—and too repeatedly—the Court had to importune and finally direct certain defense witnesses to directly answer questions posed. The trial transcripts will identify these persons as Dr. Trevor Laird and Dr. Michael Zaworotko.See, e.g., Laird Tr. 1450:20–1451:5, 1486:23–1487:10, 1488:5–22; Zaworotko Tr. 1650:16–19, 1684:5–24, 1712:18–19. Ironically, such representatives of science were more interested in evasion than intellectual candor. As to their testimony, the Court as fact finder invokes falsus in uno, falsus in omnibus. Each did the cause of his particular Defendant no good.

In challenging the patents' validity, Defendants face a heavy burden. By statute, issued patents are "presumed valid," 35 U.S.C. § 282 (2012), and Defendants must overcome that presumption by clear and convincing evidence. ActiveVideo Networks, Inc. v. Verizon Commc'ns, Inc., 694 F.3d 1312, 1327 (Fed.Cir.2012). "Clear and convincing evidence is such evidence that produces ‘an abiding conviction that the truth of [the] factual contentions are highly probable.’ " Id. (quoting Colorado v. New Mexico, 467 U.S. 310, 316, 104 S.Ct. 2433, 81 L.Ed.2d 247 (1984) ).

The Court finds that neither Defendant met its obligation to overcome the presumption of validity of the patents-in-suit by clear and convincing evidence.

I. Findings of Fact

To the extent any finding of fact reflects a legal conclusion, it shall be to that extent deemed a conclusion of law, and vice versa.

A. Background

i. The patents at issue

On October 24, 2006, the Patent and Trademark Office ("PTO") issued the ' 015 Patent, entitled "Method for the Preparation of Hexahydro–furo–[2,3–b]furan–3–ol." Pre–Trial Order at 12, ¶ 34. Lupin does not contest—for the purposes of this litigation only—that the '015 Patent has a priority date of September 10, 2001, nor that Janssen R & D Ireland holds title to the ' 015 Patent. Id. at 12, ¶¶ 35, 36. Janssen is asserting claim 1 of the ' 015 Patent against Lupin. Id. at 12, ¶ 37.

On August 10, 2010, the PTO issued the '411 Patent, entitled "Process for the Preparation of (3R,3aS,6aR)–hexahydrofuro[2,3–b]furan–3–yl(1S,2R)–3–[[ (4–aminophenyl) sulfonyl](isobutyl) amino]–l–benzyl–2–hydroxypropylcarbamate." Id. at 13, ¶ 44. Mylan does not contest—for the purposes of this litigation only—that the '411 Patent has a priority date of December 23, 2003, nor that Janssen R & D Ireland holds title to the '411 Patent. Id. at 13, ¶ 45, 46. Janssen is asserting claim 13 of the '411 Patent against Mylan. Id. at 13, ¶ 47.

On April 20, 2010, the PTO issued the '645 Patent, entitled "Pseudopolymorphic Forms of a HIV Protease Inhibitor." Id. at 11, ¶ 29. Lupin does not contest—for the purposes of this litigation only—that the '645 Patent has a priority date of May 16, 2002, nor that Janssen R & D Ireland holds title to the '645 Patent. Id. at 11, ¶¶ 30, 31. Janssen is asserting claim 4 of the '645 Patent against Lupin. Id. at 11, ¶ 32.

ii. Prezista (darunavir )

Janssen Products, L.P. is the holder of approved New Drug Application ("NDA") No. 21–976 for Prezista (darunavir ). Janssen manufactures, markets and sells Prezista, a human immunodeficiency virus (HIV–1) protease inhibitor. Id. at 9, ¶¶ 14, 15. The FDA has approved Prezista for the treatment of HIV–1 infection in adult patients and in pediatric patients 3 years of age and older. Id. at 9, 16. The active pharmaceutical ingredient in Prezista is darunavir in an ethanolate form. Id. at 9, ¶ 17. Prezista tablets are currently sold in the United States in 75 mg, 150 mg, 400 mg, 600 mg, and 800 mg dosage strengths. Id. at 9, ¶ 18. The '645 Patent is listed in an FDA publication entitled Approved Drug Products with Therapeutic Equivalence Evaluations (known as the "Orange Book") with respect to Prezista. Id. at 9, ¶ 20.

iii. The ANDAs

On June 23, 2010, Mylan filed ANDA No. 202136 with the FDA, seeking approval to sell darunavir tablets, 75 mg, 150 mg, 300 mg, 400 mg, and 600 mg, described in the ANDA. Id. at 9, 1121.

On June 23, 2010, Lupin filed ANDA No. 202073 with the FDA seeking approval to sell darunavir tablets, 400 mg and 600 mg, described in the ANDA. On June 3, 2011, Lupin amended its ANDA No. 202073 to incorporate additional strengths of 75 mg, 150 mg, and 300 mg. Id. at 10, ¶¶ 23–24.

iv. Claim construction

In its Markman opinion, the Court construed "solvate" in the asserted claims of the '645 Patent to mean "a crystal form that contains stoichiometric or non-stoichiometric amounts of solvent." ECF No. 477 at 5; Pre–Trial Order at 13, ¶ 48. The Court also construed the term "compound of formula (6)" in the asserted claims of the '411 Patent as "darunavir," or the graphic depiction of darunavir. Markman Op. at 11; Pre–Trial Order at 13–14, ¶ 49.

B. The '015 Patent

i. The invention and the claim at issue

The invention of the '015 Patent"relates to a method for the preparation of [bis-THF]...." Defendants' Trial Exhibit ("DTX") 7 ('015 Patent) at col. 1:12–14. The specification states that bis-THF "is an important pharmacological moiety present in the structure of retroviral protease inhibitors ..." id. at col. 1:18–24, and identifies and incorporates by reference the European versions of G.D. Searle & Co.'s ("Searle") darunavir patent, ¹id. at col. 1:18–24. It recommends bis-THF as having "particular interest" for use in darunavir. Id. at col. 28:24–31; col. 28:37–40.The specification also indicates that the "main object of the ... ['015] invention is to provide an improved method for producing [bis-THF] when compared to the art-known methods and their drawbacks." Id. at col. 2:25–28. "[A]nother object" of the '015 Patent is "to provide a method for the synthesis of [bis-THF] which is suitable for industrial scaling-up." Id. at col. 2:28–30.

1 After discovering that certain compounds of interest were of a family of compounds covered by patents owned by Searle, Tibotec approached Searle and obtained a license to use the compounds that Searle had discovered, including darunavir, in further research and development efforts. Stoffels Tr. 89:22–25, 90:20–91:8.

Claim 1 of the '015 Patent, which is being asserted against Lupin, recites:

A method for the synthesis of hexahydro–furo[2,3–b]furan–3–ol of formula (7) starting from an intermediate of formula (1) ... transforming said intermediate of formula (1) into a nitromethane derivative of formula (3) ... subsequently transforming said nitromethane derivative into a tetrahydrofuran derivative of formula (6) ... and then transforming the intermediate of formula (6) into hexahydro–furo[2,3–b]furan–3–ol of formula (7) by way of an intermolecular cyclisation reaction.

Id. at col. 38:44–39:40.

ii. Development of the '015 Patent invention

Tibotec, later acquired by Johnson & Johnson in 2002, Stoffels Tr. 98:18–21; 99:1–4, began considering compounds containing bis-THF for possible use as protease inhibitors, and immediately recognized the difficulty of synthesizing bis-THF. Reider Tr.2004:17–23. The bis-THF molecule has a double-ring structure, with three "chiral centers," also known as "stereocenters" or "optical centers." DTX 7 ('015 Patent) at col. 9:21–28. The three chiral centers made bis-THF an "incredibly complex" molecule that presented a "formidable challenge" to synthesize. Reider Tr.2004:17–23; see also id. 2008:23–2009:3; Ganem Tr. 1159:11–15. Tibotec initially explored whether it could avoid the problem by replacing bis-THF with a different chemical group that would have similar benefits but would be easier to synthesize. Wigerinck Tr. 186:19–188:5; see generally Plaintiff's Trial Exhibit ("PTX") 346 (Laboratory Notebook 37); PTX 351 (Tibotec Feb. 2001 Powerpoint) at PREZ04980686. Tibotec made and tested many compounds containing alternatives to bis-THF but none of the compounds had activity comparable to the compounds with bis-THF. Wigerinck Tr. 190:17–191:10, 200:12–20; PTX 351 (Tibotec Feb. 2001 Powerpoint) at PREZ04980686. After concluding that bis-THF was a "special and crucial" component for the design of protease inhibitors, Tibotec began developing a process for making bis-THF that could be used for synthesis on an industrial scale. Wigerinck Tr. 191:11–17.

Tibotec began working on this scalable synthesis for bis-THF in May 1999. Id. at 238:22–239:9. The work done over the next two years, including attempting more than twenty methods for making bis-THF before selecting the route claimed in the '015 Patent, is summarized in a progress report that Tibotec prepared in October 2000. Id. at 257:7–258:14; PTX 404 (Tibotec Progress Report). Initially, Tibotec tried the synthesis that was favored in the prior art, i.e., starting with a compound known as dihydrofuran, essentially the "left ring" of bis-THF's double ring, and proceeding from there to build the right-hand ring. Wigerinck Tr. 239:24–241:12, 258:15–259:10. This strategy was described, inter alia, in the earliest prior art synthesis of bis-THF, the Pezechk 1986 Article (PTX 391), and in multiple references by Dr. Arun Ghosh, one of the pioneers to use bis-THF as part of an HIV protease inhibitor, including Scheme 2 of the Ghosh 1996 Article (PTX 389). See Wigerinck Tr. 239:24–241:12, 243:2–7. At trial, the experts from both sides agreed that starting with bis-THFs left-hand ring was a logical strategy. See Reider Tr. 2009:21–2010:16 ("the literature taught that this is a great place to start"); Ganem Tr. 1204:2–5 (agreeing that this is a "reasonably straightforward way[ ] of thinking about how to make a two-ring bicyclic structure").

However, when Tibotec used this strategy, it encountered a problem with a step in the synthesis described in the Pezechk 1986 Article and the Ghosh 1996 Article. Both methods used ozonolysis, a reaction step using ozone, with which Tibotec and the prior art had not had a problem on a laboratory scale. Wigerinck Tr. 243:15–244:4; Ganem Tr. 1207:19–24. But Tibotec found that when it attempted the ozonolysis step on a larger scale, the reaction would generate smoke, even after stopped, which suggested that the reaction could not be controlled and had the potential to explode. Wigerinck Tr. 243:15–245:13.

In the end, none of the routes that Tibotec explored starting with bis-THF's left-hand ring proved satisfactory. See PTX 404 (Tibotec Progress Report) at PREZ05110825–26.

Tibotec also explored several routes that started with bis-THF's right-hand ring and then built the molecule's left ring. Unlike dihydrofuran, the right hand ring was not commercially available, "so that made the route more complex." Wigerinck Tr. 259:20–260:11; see PTX 404 (Tibotec Progress Report) at PREZ05110827–32. None of these efforts succeeded. Id.

Finally, Tibotec explored the use of starting materials that had neither of bis-THF's rings and thus entailed creating both rings. Wigerinck Tr. 260:14–25; PTX 404 (Tibotec Progress Report) at PREZ05110833–42. The scientists at Tibotec, including Dr. Wigerinck, "did not expect [this concept] to work." Wigerinck Tr. 260:14–25. Creating both rings from scratch was the approach that Dr. Ghosh had used in his initial synthesis of bis-THF for Merck in the early 1990s—an approach that Merck had rejected as unduly complicated: Dr. Ghosh believed "nobody in [his] right mind would launch a project based upon this literature of bis-tetrahydrofuran, when they see the synthesis [is] so complex." Tr. 2457:24–2458:2. Tibotec failed in its efforts to reproduce the process that Dr. Ghosh had used, Wigerinck Tr. 241:24–242:8, but developed different routes which also involved creating both rings from scratch.

The route that Tibotec ultimately adopted and claimed in the '015 Patent was described in the October 2000 Progress Report as Route 4.1. PTX 404 (Tibotec Progress Report) at PREZ05110833–39; Ganem Tr. 1211:2–8. This route starts with an "open chain molecule," Reider Tr. 2011:14–17, with chirality built in. The starting material is a "protected glyceraldehyde" known as glyceraldehyde acetonide, and Tibotec referred to this route internally as the "glyceraldehyde acetonide" or "protected glyceraldehyde" route. PTX 404 (Tibotec Progress Report) at PREZ05110833; Wigerinck Tr. 260:14–261:4. The protected glyceraldehyde is referred to in the '015 Patent as compound (1). See DTX 7 ('015 Patent) at col. 2:50–65.

Tibotec recognized that the protected glyceraldehyde route was problematic in several respects. Glyceraldehyde acetonide was not commercially available; "you can't buy it." Wigerinck Tr. 202:20–203:17. It was also unstable and could not be stored for long periods of time. Id.; Ganem Tr. 1211:9–1212:2. In addition, glyceraldehyde acetonide "has the chirality in it from the very beginning at the center," which complicates the synthesis. Reider Tr. 2012:8–24.Despite the complexity of the protected glyceraldehyde route, Tibotec came to recognize that it was "the best route from both an economic and a technical viewpoint." PTX 396 (E-mail from M. Lubben dated May 22, 2001) at PREZ05001242–43; Wigerinck Tr. 261:24–262:25. Tibotec selected this route for its commercial process and uses it to this day. Reider Tr. 2025:1–15. The "protected glyceraldehyde route" is described and claimed in the '015 Patent, and it is the subject of claim 1 of the '015 Patent.

Lupin uses the claim 1 method to make bis-THF for use as a component of its ANDA products. This Court has granted summary judgment that Lupin's importation or sale of darunavir containing bis-THF made by that method would infringe claim 1 of the '015 Patent. Summ. J. Op. at 49–59 (ECF No. 769).

C. The '411 Patent

i. The invention and the claim at issue

The '411 Patent has a priority date of December 23, 2003, the latest of the three patents at issue in this litigation. Pre–Trial Order at 13; DTX 13 ('411 Patent) (cover). The '411 Patent discloses the full "suite" of Tibotec's darunavir patents, "bring[ing] all the pieces together" and enabling production of the darunavir molecule as a pharmaceutically acceptable drug capable of being manufactured on an industrial scale. Reider Tr. 1990:22–1991:19, 2035:5–12.

Asserted claim 13 of the '411 Patent depends from claim 2, which in turn depends from claim 1. The claim requires a five-step process for synthesizing darunavir. Independent claim 1 recites a four-step synthesis of the right-hand sulfonamide component of the darunavir molecule (referred to in the patent as the "compound of formula (5)"). The third step is "(iii) reducing the nitro moiety of the resultant compound of step (ii)," and the separate fourth step is "(iv) deprotecting the resultant compound of step (iii)." DTX 13 ('411 Patent) at col. 23:45–48. The fifth and final step involves coupling the compound of formula (5) to a bis-THF derivative to form darunavir. Id. at col. 23:48–51; Reider Tr. 2035:20–2036:5; Laird Tr. 1460:23–1461:8.

Claim 2, which depends from claim 1, adds a requirement for a "tert-butoxycarbonyl," or "boc" protecting group to be used in the chemical intermediates of claim l.² DTX 13 ('411 Patent) at col. 23:52–25:3; Reider Tr. 2051:5–25. The boc protecting group is one of many possible protecting groups, DTX 13 ('411 Patent) at col. 6:27–67, and before the '411 Patent, had never been used in the synthesis of darunavir, Reider Tr. 2051:7–12.

2 A protecting group protects part of the molecule from unwanted reactions while other parts of the molecule are being worked on. Reider Tr. 2038:20–2039:10.

Claim 13, which depends from claim 2, identifies a specific coupling agent, BNPC, for use in the final step of the process in which the bis-THF moiety is coupled with the sulfonamide moiety to form darunavir. Id. at 2054:4–9; DTX 13 ('411 Patent) at col. 26:10–13.

ii. Development of the '411 Patent invention

The invention of the '411 Patent is a method for making darunavir by making the sulfonamide, or "right-hand side" of the darunavir molecule, and then coupling the sulfonamide with the left-hand bis-THF to form the completed darunavir molecule. Reider Tr. 1991:15–19. The inventors of the '411 Patent did not set out to create a new manufacturing process for darunavir. Wigerinck Tr. 176:19–177:8. Rather, they expected to use the process developed by Searle that Searle had granted Tibotec a license to use under Searle's '775 Patent, which discloses a commercially suitable method of manufacturing darunavir. Id.; Stoffels Tr. 90:23–91:8. Searle had successfully manufactured approximately 100 kilograms of a starting material suitable to make the right-hand sulfonamide component of darunavir. They transferred that material to Tibotec where the Tibotec inventors initially worked with it. Wigerinck Tr. 281:5–282:2. This starting material had a "dibenzyl" protecting group. Id.

Tibotec recognized that Searle's dibenzyl-protected starting material provided an "attractive" and "short synthesis route" for making darunavir. Id. at 281:5–282:2. Specifically, the Searle dibenzyl route allowed two chemical transformations—the "reduction" and "deprotection" transformations—to be consolidated into a single step, thus saving time and expense. Id. at 282:22–283:6. However, Tibotec "unexpectedly" experienced "big, big problems" with the Searle dibenzyl route. Id. at 282:3–16. The route had a tendency to "poison" the palladium catalyst used in the hydrogenation step used for deprotecting, or removing, the dibenzyl protecting group. Id. at 282:3–18, 284:9–285:17. Catalyst poisoning refers to the tendency of certain contaminants to bind to the catalyst instead of the desired chemical compound. Reider Tr. 2046:1–16. This slows down the reaction, leading to high levels of impurities and the need for excessive amounts of the palladium catalyst, a precious metal "more expensive than gold." Wigerinck Tr. 283:23–284:8, 285:10–17; PTX 374 (ChemShop Project Report) at PREZ04997815.

In addition to catalyst poisoning, Tibotec experienced safety issues with the Searle dibenzyl route, specifically serious exothermic reactions that suddenly released heat and had the potential to cause a "runaway" reaction and explosions. Wigerinck Tr. 283:7–20.

The Searle '775 Patent also taught the use of a different protecting group, the benzyloxy carbonate, or CBZ, protecting group. Reider Tr. 2039:17–21. Tibotec considered using CBZ instead of dibenzyl, but did not pursue this option because the two protecting groups were removed "in an identical way" and Tibotec "would be facing the same difficulties" as it had with the dibenzyl-protected process. Wigerinck Tr. 287:14–19.

Eventually Tibotec concluded that it could solve the problems it had identified only by giving up the major advantage of the Searle route, which was performing reduction and deprotection simultaneously in one step. Tibotec separated these into two separate steps, which increased cost and lengthened the time of production. Id. at 282:22–283:6. Tibotec accomplished the separation of the deprotection and reduction steps by replacing the dibenzyl protected starting material with a "boc" protected starting material. Id. at 285:18–286:2. Unlike the dibenzyl group, the boc protecting group did not require a catalyzed hydrogenation reaction for deprotection, and as a result, Tibotec could better control the reaction that had previously led to an exotherm and poisoned catalyst. Id.

In addition to building the sulfonamide, the '411 Patent covers the final process step in which bis-THF is attached to the rest of the darunavir molecule—the "coupling" step. Tibotec also sought to improve this step. As discussed, bis-THF is separately manufactured by the process claimed in the ' 015 Patent. DTX 13 ('411 Patent) at col. 15:4–6. Then it is coupled to the sulfonamide, or right-hand side of the molecule, which is called the compound of formula (5) in the '411 Patent. Id. at col. 15:6–10; 16:26–48. This coupling step requires a coupling agent, which activates the bis-THF so that the coupling reaction can take place. Reider Tr. 2035:24–2036:12. Tibotec first tried disuccinimidyl carbonate ("DSC") as a coupling agent, but was dissatisfied and searched for years for a better option; it even hired an independent laboratory, Carbogen, to study alternatives to DSC. Wigerinck Tr. 295:5–21; PTX 434 (Carbogen Process Research Report) at PREZ5073327. Carbogen concluded that "commercially available alternative coupling agents with promising reactivity were not uncovered." PTX 434 (Carbogen Process Research Report) at PREZ5073327.

Tibotec continued to experiment, and two years later, discovered the bis–(4–nitrophenyl) carbonate, or BNPC, coupling agent recited in claim 13 of the ' 411 Patent. PTX 433 (Chem–Pharm Team Meeting Minutes) at PREZ05041021. Tibotec claimed the BNPC coupling agent in the '411 Patent. DTX 13 ('411 Patent) at col. 26:10–13.

Before trial, the Court granted summary judgment to Janssen that Mylan infringes claim 13 of the '411 Patent. Summ. J. Op. at 5–8.

D. The '645 Patent

i. The invention and the claim at issue

The invention of the '645 Patent is the ethanolate form of the drug that Janssen developed and sells as Prezista. Asserted claim 4 is a dependent claim that depends from claim 3. Claim 3 recites "[a] composition comprising an ethanolate solvate of the compound [darunavir ], in which the ratio of compound to ethanol is about 1:1, and an inert carrier." PTX 1 ('645 Patent) at col. 30:35–40. Claim 4 recites "[t]he composition of claim 3 wherein the inert carrier is a pharmaceutically acceptable carrier." Id. at col. 30:41–42.

ii. Development of the '645 Patent invention

When Tibotec initially synthesized darunavir, the compound was in the form of a "sticky oil," which Tibotec could not use for development or testing. Wigerinck Tr. 303:16–22; 304:19–305:1. Tibotec set out to transform the "sticky oil" formulation of darunavir into a solid powder form. Id. at 299:24–300:5, 303:16–22. Initially, Tibotec tried making salt forms of darunavir, as salts are common for pharmaceuticals (approximately half of all marketed pharmaceuticals are salts) since they typically have good solubility, which allows them to dissolve in the body.Id. at 300:18–301:8; Myerson Tr. 638:18–639:6; Gould Tr. 1291:4–6, 1291:23–24. Tibotec created and tested multiple salt forms, and while the salt forms of darunavir could have been used and delivered inside a capsule, Tibotec determined that they were not optimal. Wigerinck Tr. 301:9–16, 335:9–11; Reider Tr. 2080:19–2081:5.

Tibotec also made prodrugs of darunavir, Wigerinck Tr. 300:18–21, 301:17–202:15, but rejected that formulation because it was "quite complex to make." Id. at 303:11–15.

After rejecting both salts and prodrug formulations of darunavir, Tibotec sought a different formulation, and to assist in this effort it hired ChemShop, a laboratory which had expertise in making solid forms of pharmaceutical substances. Id. at 284:15–17, 303:16–24, 306:14–24. ChemShop found that making a solid powder form of darunavir was "unexpectedly difficult," id. at 303:16–24, and did not succeed in crystallizing darunavir from any solvent, id. at 306:14–24. After ChemShop's failure, Dr. Wigerinck brought the project back in-house at Tibotec where the Tibotec chemists repeated the crystallization experiments that ChemShop had performed, but again failed. Id. at 303:24–304:4. After that, over the course of several weeks, Dr. Wigerinck had each of Tibotec's 20 chemists perform two crystallization experiments a day, one in the morning and one in the afternoon, so that in total Tibotec was performing "more than 40 experiments a day, just to try to convert the sticky oil to a white powder." Id. at 304:5–11. In these experiments, Tibotec used many different solvents, including ethanol, and a variety of other conditions. Id. at 307:12–18, 311:9–16. Tibotec failed to obtain a crystalline powder form of darunavir from ethanol. Id. at 311:9–13.

When these efforts failed, Dr. Wigerinck instructed his scientists to employ a "scratching" technique to try to "obtain a first crystal" of darunavir. Id. at 304:5–18, 307:22–25. The technique is "long-known" in the art for trying to create a crystalline form of a compound. Zaworotko Tr. 1614:9–13. In these experiments, Tibotec continued to vary the solvents, anti-solvents, temperatures ("25 degrees, 60 degrees, 80 degrees"), purification methods, and even the rates of scratching. Wigerinck Tr. 319:1–322:12. Eventually, using isopropanol as a solvent, Tibotec discovered conditions that would make a powder—a relatively pure crystalline form of darunavir. Id. at 323:21–325:1; PTX 344 (Vergouwen Laboratory Notebook) at PREZ0007450; DTX 2088 (Translation of PTX 344); Wigerinck Tr. 532:10–533:19, 536:7–16.

This crystalline material that Tibotec had made turned out to be a solvate—a crystal in which the solvent is part of the crystal. Myerson Tr. 570:23–571:3. This was not what Tibotec had hoped for, as solvates are "not that easy to handle" and "are not ideal because they can have limited stability" Wigerinck Tr. 178:11 –13, 299:21 –23, 332:14–15, 494:19–21.

The solvate from isopropanol was also problematic because there were unacceptable amounts of isopropanol sticking to the powder. Id. at 325:15–326:5. Using an isopropanol solvate can present a "toxicity issue" and the amount of isopropanol in the resulting solvate was "above what a person [should] ingest." Myerson Tr. 589:4–13. To solve this problem, Tibotec tried recrystallizing the isopropanol solvate in an ethanol/water mixture, but the results were still unsatisfactory. Tibotec then repeated the dissolution and re-crystallization steps a second time on the resulting powder, this time using pure ethanol as a solvent. Wigerinck Tr. 324:22–328:1; PTX 377 (Tibotec Research Information Report) at PREZ05002056. This entire process took Tibotec six weeks. Wigerinck Tr. 326:6–328:1.

Using this three-step crystallization procedure, Tibotec was able to make an ethanolate where the ratio of darunavir to ethanol in its ethanolate was about 1:1. PTX 377 (Tibotec Research Information Report) at PREZ05002056; Wigerinck Tr. 332:7–18, 327:1–328:1. The three-step process is described in the specification of the '645 Patent as Example 1. DTX 1 ('645 Patent) at col. 15:49–67; Myerson Tr. 588:4–590:9; Zaworotko Tr. 1713:22–1714:3.

After creating an ethanolate of darunavir, Tibotec continued to search for a solid form of darunavir that would not be a solvate. Wigerinck Tr. 332:7–333:16. Tibotec commissioned a study by a third-party contractor to "obtain [darunavir ] in a non solvated form," but its efforts were unsuccessful. PTX 356 (Solvias Polymorphism Study Report) at PREZ0497447–8; Wigerinck Tr. 333:3–16, 334:25–335:11. Tibotec was never able to find a better formulation than darunavir ethanolate. Wigerinck Tr. 335:9–11.

Before trial, Lupin entered into a so-ordered stipulation stating that it does not dispute infringement of the asserted claims of the '645 Patent (including claim 4) and that "a judgment of infringement will be entered against Lupin" as to "any of the Asserted Claims [that] are not invalid...." ECF No. 375.

II. Conclusions of Law

A. Legal Standards for Obviousness, Enablement and Written Description Under the patent statute, a "patent is presumed to be valid, 35 U.S.C. § 282 (1994), and this presumption can only be overcome by clear and convincing evidence to the contrary." Bristol–Myers Squibb Co. v. Ben Venue Labs., Inc., 246 F.3d 1368, 1374 (Fed.Cir.2001).

"[O]bviousness is ultimately a question of law...." Boston Scientific Scimed, Inc. v. Cordis Corp., 554 F.3d 982, 990 (Fed.Cir.2009) ; 35 U.S.C. § 103. That "legal question" is "based on underlying factual determinations," Unigene Labs., Inc. v. Apotex, 655 F.3d 1352, 1360 (Fed.Cir.2011), which include: "1) the scope and content of the prior art; 2) the level of ordinary skill in the art; 3) the differences between the claimed invention and the prior art; and 4) evidence of secondary factors, also known as objective indicia of nonobviousness." Id. (listing what are called the "Graham factors" from Graham v. John Deere Co., 383 U.S. 1, 17–18, 86 S.Ct. 684, 15 L.Ed.2d 545 (1966) ). "Obviousness requires more than a mere showing that the prior art includes separate references covering each separate limitation in a claim under examination." Id. (citing KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398, 418, 127 S.Ct. 1727, 167 L.Ed.2d 705 (2007) ). "Rather, obviousness requires the additional showing that a person of ordinary skill at the time of the invention would have selected and combined those prior art elements in the normal course of research and development to yield the claimed invention." Id. Obviousness is judged from the perspective of a hypothetical person of ordinary skill in the art, who is "a person of ordinary creativity, not an automaton." KSR, 550 U.S. at 421, 127 S.Ct. 1727.

In addition, an "obviousness determination requires that a skilled artisan would have perceived a reasonable expectation of success in making the invention in light of the prior art." In re Cyclobenzaprine Hydrochloride Extended–Release Capsule Patent Litig., 676 F.3d 1063, 1069 (Fed.Cir.2012) (citation omitted). Courts must be aware of the "danger of hindsight bias" in an obviousness analysis. Iron Grip Barbell Co. v. USA Sports, Inc., 392 F.3d 1317, 1320 (Fed.Cir.2004) ; see also KSR, 550 U.S. at 421, 127 S.Ct. 1727 ; Graham, 383 U.S. at 36, 86 S.Ct. 684 ; In re NTP, Inc., 654 F.3d 1279, 1299 (Fed.Cir.2011).

Challenges to validity based on inadequate written description and enablement require resolution of facts. See Trading Techs. Int'l, Inc. v. Open E Cry, LLC, 728 F.3d 1309, 1318 (Fed.Cir.2013) ; Wyeth & Cordis Corp. v. Abbott Labs., 720 F.3d 1380, 1384 (Fed.Cir.2013). The written description requirement is satisfied if the patent specification "clearly allow [s] a person of ordinary skill in the art to recognize that [the inventor] invented what is claimed." Crown Packaging Tech., Inc. v. Ball Metal Beverage Container Corp., 635 F.3d 1373, 1380 (Fed.Cir.2011). The enablement requirement "is met when at the time of filing the application one skilled in the art, having read the specification, could practice the invention without ‘undue experimentation.’ " Cephalon, Inc. v. Watson Pharms., Inc., 707 F.3d 1330, 1336 (Fed.Cir.2013) (citation omitted). To satisfy the enablement requirement a patent does not need to describe what was known in the art. Epistar Corp. v. ITC, 566 F.3d 1321, 1336–37 (Fed.Cir.2009).B. The '015 Patent

i. Obviousness

1. Level of Ordinary Skill in the Art

Janssen and Lupin offered separate definitions for a person having ordinary skill in the art. Dr. Paul Reider, on behalf of Janssen, said that a person of ordinary skill in the art would have had a master's degree in chemistry and several years of laboratory experience. Reider Tr. 2003:4–7. Dr. Bruce Ganem, on behalf of Lupin, opined that a person of ordinary skill in the art would have a Ph.D. in chemistry and work experience in the pharmaceutical industry specifically directed to synthetic chemistry. Ganem Tr. 1001:10–16, 1003:15–1004:12. The Court finds that either of these definitions is satisfactory and that regardless of which is chosen, the analysis herein would be the same. Dr. Reider and Dr. Ganem each stated that their opinions on validity did not depend on the definition of the level of ordinary skill in the art and would be the same under either side's definition. Reider Tr. 2003:17–20; Ganem Tr. 1006:11–14. The Court adopts Dr. Reider's broader definition of a person of ordinary skill in the art which the trial evidence has shown to be pragmatically realistic.

2. Scope and Content of the Prior Art

Under Graham, the Court must define the scope and content of the prior art as of September 10, 2001, the priority date of the ' 015 Patent. Pre–Trial Order ¶ 35. A reference qualifies as prior art for an obviousness determination under 35 U.S.C. § 103 only when it is analogous to the claimed invention. Innovention Toys, LLC v. MGA Entm't, Inc., 637 F.3d 1314, 1321 (Fed.Cir.2011). "Two separate tests define the scope of analogous prior art: (1) whether the art is from the same field of endeavor, regardless of the problem addressed and, (2) if the reference is not within the field of the inventor's endeavor, whether the reference still is reasonably pertinent to the particular problem with which the inventor is involved." In re Bigio, 381 F.3d 1320, 1325 (Fed.Cir.2004). "A reference is reasonably pertinent if, even though it may be in a different field from that of the inventor's endeavor, it is one which, because of the matter with which it deals, logically would have commended itself to an inventor's attention in considering the problem." In re Clay, 966 F.2d 656, 659 (Fed.Cir.1992).

Janssen identifies seven prior art references that describe methods of making bis-THF:

The Pezechk 1986 Article (PTX 391): This article describes a method for synthesizing bis-THF that starts with dihydrofuran, an achiral cyclic compound virtually identical to the lefthand ring in bis-THF's double-ring structure. PTX 391 at PREZ05017244–45.

The Ghosh 1994 Article (DTX 91): This article "touts bis-THF" for use as a component for analogs of a protease inhibitor known as saquinavir. Ganem Tr. 1217:15–18, 1196:4–8; DTX 91 at PREZ05005264. The method described, developed by Merck, starts with diethyl malate, a linear, open chain (non-ring) structure with a chiral center. DTX 91 at PREZ05005263; Ganem Tr. 1159:16–22.

The Ghosh 1995 Article (PTX 222): In this 1995 article, Dr. Ghosh adopted the approach that Pezechk advocated earlier and synthesized bis-THF by starting with dihydrofuran and constructing only a second ring. PTX 222 at 507; Ganem Tr. 1117:11–16.

The Ghosh 1996 Article (PTX 389): This article describes two methods for making bis-THF. Scheme 1 is the diethyl malate route previously disclosed in the Ghosh 1994 Article. PTX 389 at PREZ04989981; Ganem Tr. 1159:16–22. Scheme 2 starts with dihydrofuran and is similar to the route described in the Ghosh 1995 Article. PTX 389 at PREZ04989981.

U.S. Patent No. 7,470,506 (PTX 35) : The '506 Patent was filed in 1998. Dr. Erickson is the first named inventor and Dr. Ghosh is one of three co-inventors. Like the Pezechk 1996 Article and various articles by Dr. Ghosh, the '506 Patent teaches a synthesis that starts with dihydrofuran. PTX 35 at Fig. 2.

The Ghosh 1999 Article (PTX 392): This 1999 article also uses dihydrofuran as a starting point, with certain improvements in the method described in Dr. Ghosh's earlier articles. PTX 392 at PREZ05142935.

The Uchiyama 2001 Article (DTX 790): The route described in this article also starts with dihydrofuran. DTX 790 at 4654–55.

Based on these references, Janssen argues that the prior art taught that a method using bis-THF's left-hand ring as a starting point was desirable, see PTX 222 (Ghosh 1995 Article) at 506 (describing the method as "convenient," "efficient," and "operationally simple"), and that nothing in any prior art reference suggested that this method had any drawbacks or disadvantages. Janssen asserts that the method of claim 1 of the '015 Patent is not identified anywhere in the bis-THF prior art, and that to find prior art that describes these steps of the '015 Patent's method (involving a protected glyceraldehyde, nitromethane, and the use of a Michael addition reaction and a Nef reaction—reactions used in the claimed method of the '015 Patent ), Lupin had to go outside the relevant art and turn to references that do not relate to protease inhibitors in general or to bis-THF in particular.

Lupin argues, however, that a person of ordinary skill in the art would have been motivated to devise an improved synthesis of bis-THF in view of the known drawbacks of prior art syntheses, and that a person of ordinary skill in the art would have been motivated to apply retrosynthetic analysis in order to find an improved synthesis route. Retrosynthetic analysis is a technique originally developed by Nobel laureate Dr. Elias Corey. Ganem Tr. 1011:10–13; see generally DTX 1400 (Corey 1967); DTX 1399 (Corey 1975). The technique involves developing a synthetic scheme by working backwards from a target molecule and in an iterative manner breaking strategic bonds in order to identify a simpler compound. The goal of the technique is to get to a starting material that is commercially available or could be easily prepared from the literature along with a set of well-known or otherwise established reactions that then may be used in the forward direction to synthesize the target molecule. Ganem Tr. 1009:4–12, 1013:22–1014:5; Reider Tr. 2138:24–2139:15; DTX 1399 (Corey 1975) at 6117; DTX 1400 (Corey 1967) at 19–21. According to Lupin, a person of ordinary skill in the art applying retrosynthetic analysis would have been led to prior art identifying key starting materials, intermediates and the Michael addition reaction using nitromethane suitable in devising a bis-THF synthetic scheme. In particular, Lupin identifies four references that it argues should be considered prior art, summarized below:

The Patrocinio 1994 and Costa 1997 Articles (DTX 792A and DTX 793A): These articles are two publications from the same laboratory that describe the same study, with Costa providing additional detail. Reider Tr. 2022:2–13. These references identify a protected glyceraldehyde (a synthetically useful chiral building block easily obtained from inexpensive D (+)-mannitol ) and the use of a Michael addition reaction using nitromethane as a reagent for the transformation of compound (1) to compound (3). Ganem Tr. 1087:2–1088:12.

The Seebach 1979 and Rosini 1988 Articles (DTX 794A and DTX 795A): These articles, specifically referred to in the Patrocinio 1994 and Costa 1997 Articles, describe chemistry performed on a class of compounds known as nitroalkanes. They disclose the Nef reaction ("the transformation of a nitroalkene derivative into a carbonyl derivative," Kesteleyn Tr. 984:4–10), one of the reactions used in the claimed method of the '015 Patent. Ganem Tr. 1065:23–1067:19.

The Patrocinio 1994, Costa 1997, Seebach 1979, and Rosini 1988 Articles do not mention protease inhibitors in general or bis-THF in particular. Reider Tr. 2022:14–17, 2023:19–22; Ganem Tr. 1191:5–8. The ideas in those references were well known in the art between 1994 and 2001 when the relevant bis-THF art was being developed, but no one in the art turned to their teachings for the synthesis of bis-THF. Ganem Tr. 1070:7–21, 1072:6–12, 1118:25–1119:21. When asked by this Court why no one looked to the Patrocinio 1994, Costa 1997, Seebach 1979, and Rosini 1988 Articles in designing routes for synthesizing bis-THF if it was obvious to do so, Dr. Ganem only stated: "It's possible that they didn't search for it. I don't know." Ganem Tr. 1070:14–18. The Court finds that this testimony demonstrates that it is only with hindsight that one of ordinary skill in the art would have come across these four references.

These four references are not analogous prior art because they would not logically have commended themselves to a person skilled in the art. This Court is not convinced that the retrosynthetic analysis described by Dr. Ganem "unambiguously" leads to these four references, and subsequently to the starting point of the claim 1 synthesis. Ganem Tr. 1186:16–22. Retrosynthetic analysis is a "stimulating tool for creativity" that allows a chemist to consider a wide array of possible synthetic routes. Reider Tr. 2019:21–2020:11. As credibly explained by Dr. Reider, referencing the work by the pioneer of retrosynthetic analysis Dr. Corey, retrosynthetic analysis does not lead unambiguously anywhere, but rather suggests exploration and analysis of a large number of possible routes for creating molecules.Id. Dr. Ganem agreed on cross-examination that "retrosynthetic analysis doesn't point to one unequivocal choice." Ganem Tr. 1182:25–1183:3.

Dr. Ganem's retrosynthetic analysis also improperly relies on hindsight. Real-world evidence demonstrates that retrosynthetic analysis would lead persons of skill in the art to choose dihydrofuran, not a protected glyceraldehyde, as the starting point for synthesizing bis-THF. As Dr. Ganem conceded, the authors of Uchiyama 2001 applied retrosynthetic analysis in developing their synthesis of bis-THF. Id. at 1188:17–1189:22. Their application of retrosynthetic analysis resulted in the selection of dihydrofuran as a starting point. Id.; DTX 790 (Uchiyama 2001 Article) at 4654. Similarly, Tibotec utilized a form of retrosynthetic analysis by studying various ways of breaking the bonds of bis-THF to arrive at different synthetic routes, and that analysis led to the dihydrofuran method taught in the prior art. Reider Tr. 2010:21–2011:17; PTX 404 (Tibotec Progress Report).

In sum, retrosynthetic analysis provides no reason to search out or come across the Patrocinio 1994, Costa 1997, Seebach 1979, or Rosini 1988 Articles. They are random references that have nothing to do with bis-THF, darunavir, protease inhibitors or HIV/AIDS drugs. As Dr. Ganem conceded, these references were available before Ghosh's work, but neither Dr. Ghosh—who had studied with Dr. Corey, the pioneer of retrosynthetic analysis, at Harvard—nor anyone else in the art used those references in a method for making bis-THF. Ganem Tr. 1070:10–18, 1118:25–1119:18. The only explanation for pulling these references out of the vast sea of chemistry literature is the knowledge—obtained from working backwards from the '015 Patent —that they can be tacked together to re-create the process that Tibotec created. As a result, the scope and content of the prior art is limited to the seven references that Janssen identifies.

3. Differences between the Asserted Claims and the Prior Art

This Court finds that the process described in claim 1 for making bis-THF is unlike the processes for making bis-THF described in the prior art. The process of claim 1 starts with a protected glyceraldehyde (referred to in the ' 015 Patent as compound (1)), which does not contain either of bis-THF's rings. That compound is then transformed using nitromethane to a compound referred to as compound (3). Compound (3) is further transformed, in part by the use of a Nef reaction, to a compound referred to as compound (6). In the final step of the claimed process, compound (6) is transformed via an intramolecular cyclization reaction to form the second ring of bis-THF. The result is compound (7), or bis-THF. See Reider Tr. 2018:6–2019:2. The claim 1 process is "counterintuitive" and "dramatically" different from the prior art methods. Reider Tr. 2013:24–2014:1, 2025:1–5. This Court finds that claim 1 would not have been obvious to a person of ordinary skill in the art.

Using a protected glyceraldehyde (compound (1) of the '015 Patent ) as the starting point for synthesizing bis-THF was "counterintuitive" and not obvious. Reider Tr. 2013:22–2014:21. Unlike the routes described in the prior art that start with a molecule containing bis-THF's lefthand ring, the claim 1 process starts with a compound that contains neither of the bis-THF rings, which means that both of the bis-THF rings have to be created, instead of only one. See Reider Tr. 2009:21–2010:16; Ganem Tr. 1117:17–24. There were also other drawbacks to starting with a protected glyceraldehyde—it is "not commercially available in scale. It's not stable, and it has chirality in it from the very beginning at the center." Reider Tr. 2012:13–24. For these reasons, a person of ordinary skill in the art would not have chosen a protected glyceraldehyde as the starting point for making bis-THF. Reider Tr. 2019:10–13. There would have been no reason, other than hindsight, for a person of ordinary skill to pick the Patrocinio and Costa references from the large library of chemical literature and to select a protected glyceraldehyde as a starting point for making bis-THF.

To resay, Dr. Ganem's retrosynthetic analysis is not persuasive. Real-world evidence, rather than conjecture in court testimony, demonstrates that restrosynthetic analysis would lead persons of ordinary skill in the art to choose dihydrofuran, not a protected glyceraldehyde, as the starting point for synthesizing bis-THF. As Dr. Ganem conceded, the authors of the Uchiyama 2001 Article applied retrosynthetic analysis in developing their synthesis of bis-THF, Ganem Tr. 1188:17–1189:22, and their application resulted in the selection of dihydrofuran as a starting point. Id.; DTX 790 (Uchiyama 2001 Article) at 4654. Similarly, in designing its program to synthesize bis-THF, Tibotec utilized a form of retrosynthetic analysis—that is, studying various ways of breaking the bonds of bis- THF to arrive at different synthetic routes. Again, that analysis led to the dihydrofuran method taught in the prior art, and Tibotec worked with that method before developing the method claimed in the '015 Patent. Reider Tr. 2010:21–2011:17; PTX 404 (Tibotec Progress Report).

Lupin attempts to turn the scope of the claim into an issue in its obviousness case by pointing out that claim 1 is not limited to an industrial scale synthesis. The '015 Patent's value is that it allows the manufacture of bis-THF on an industrial scale, but the full scope of the claim is very different from anything in the prior art, no matter how it is analyzed. The prior art taught starting with dihydrofuran and then making bis-THF's other ring and it discouraged making both rings from scratch. The prior art taught starting with a commercially available, stable compound that was achiral. This is true whether on a laboratory scale or an industrial scale. And the prior art provided no motivation to change the known methods, whether for laboratory scale or industrial scale synthesis. Dr. Ganem admitted that "nothing in the prior art stated that there was any problem with the Ghosh synthesis based on ozonolysis," at "any scale." Ganem Tr. 1207:19–24. It follows that one of skill in the art would have seen no reason to modify the Ghosh dihydrofuran route, whether for laboratory or industrial use.

4. Secondary Considerations

The last of the Graham factors directs the Court's attention to secondary considerations, or "objective indicia of nonobviousness." Graham, 383 U.S. at 17–18, 86 S.Ct. 684. "Objective indicia of nonobviousness play a critical role in the obviousness analysis." Leo Pharm. Prods., Ltd. v. Rea, 726 F.3d 1346, 1358 (Fed.Cir.2013). "Objective indicia ‘can be the most probative evidence of nonobviousness in the record, and enables the court to avert the trap of hindsight.’ " Id. (citation omitted).

a. Failure of others

Janssen first asserts that the failure of others in finding an improved synthesis route for bis-THF, particularly one that could be used on an industrial scale, goes to the nonobviousness of the '015 Patent. Janssen states that Dr. Ghosh, a chemist of "greater than ordinary skill," Ganem Tr. 1201:20–22, was unable to create a commercially viable synthesis. Dr. Ghosh described his initial method, which he worked on while at Merck and which is discussed in the Ghosh 1994 Article (DTX 91), as "not a very good step." Ghosh Tr. 2457:21–22. He said that "nobody in [his] right mind would launch a project based upon this literature ... when they see the synthesis [is] so complex." Id. at 2457:24–2458:2.

Janssen states that over the next seven years, Dr. Ghosh and others attempted to develop improved processes for making bis-THF. Dr. Ghosh's work is described in the Ghosh 1995 Article (PTX 222), the Ghosh 1996 Article (PTX 389), the '506 Patent (PTX 35), and the Ghosh 1999 Article (PTX 392). In 2001, Dr. Uchiyama also tried to improve the synthesis. DTX 790 (Uchiyama 2001 Article). Tibotec determined that the method described in these references was useful only on a laboratory scale, DTX 7 ('015 Patent) at col. 2:22–24, and that the batting average for these prior art attempts to improve the synthesis was "pretty pathetic." Ganem Tr. 1197:23–1198:2, 1201:7–10, 1217:18–21. Janssen claims that the Tibotec inventors succeeded where others had failed by developing a novel method of manufacturing bis-THF that is suitable for synthesis on both a laboratory and on an industrial scale. Reider Tr. 2009:4–8; Wigerinck Tr. 266:10–12.Lupin responds that none of the prior art syntheses of bis-THF reflect a "failure of others," but rather reflect a successful effort on the part of the researchers to synthesize bis-THF, albeit on a small scale. Ganem Tr. 1104:5–11.

This Court finds that the "failure of others" to create a commercially available synthesis of bis-THF weighs against a finding of obviousness as to claim 1 of the '015 Patent. It is clear to the Court that in the time period between 1994, when Ghosh published the synthesis route he had worked on at Merck, and 2001 when Tibotec discovered the patented process, there indeed was a failure of others to synthesize bis-THF on a commercial scale. See generally PTX 222 (Ghosh 1995 Article); PTX 389 (Ghosh 1996 Article); PTX 35 ('506 Patent); PTX 392 (Ghosh 1999 Article). The Court finds Lupin's disingenuous argument about these prior art references, which were obviously working toward a synthesis that could actually be used in commercial production and subsequently in a pharmaceutical product, rather than only on a laboratory scale, utterly unpersuasive.

b. Commercial success

Next, Janssen argues that the commercial success of Prezista is a secondary factor that weighs against a finding of obviousness. Janssen asserts that bis-THF is a key component of the darunavir molecule in Prezista, Wigerinck Tr. 191:11–17; Reider Tr. 2025:8–13, and that since its introduction, Prezista has been a huge success, with approximately one million prescriptions annually. Stoffels Tr. 104:12–16, 105:18–20, 108:13–17, 112:1–11. Worldwide sales of Prezista in 2013 were more than $1.6 billion and U.S. sales were more than $800 million. Leffler Tr. 2288:21–24, 2311:9–15; Stoffels Tr. 112:12–14. Prezista is "the number-one selling protease inhibitor." Leffler Tr. 2311:22–2312:3. Sales figures and market data are strong evidence of commercial success. See Tec Air, Inc. v. Denso Mfg. Mich., Inc., 192 F.3d 1353, 1360–61 (Fed.Cir.1999).

Lupin responds by arguing that Janssen's general allegations of commercial success lack a specific nexus to the process steps required by claim 1 of the ' 015 Patent. See In re Huai–Hung Kao, 639 F.3d 1057, 1068 (Fed.Cir.2011) ("Where the offered secondary consideration actually results from something other than what is both claimed and novel in the claim, there is no nexus to the merits of the claimed invention."). Lupin asserts that when it comes to why doctors prescribe Prezista, the particular process by which darunavir is made, including the process used to make bis-THF, has never been shown to be clinically relevant. Zingman Tr. at 861:16–25. In support, Lupin asserts that Janssen has never marketed that Prezista includes a bis-THF group, Falcon Tr. 2494:15–17, and also makes clear that bis-THF alone has weak pharmacologic activity and that it has not been approved by the FDA for any clinical use. Wigerinck Tr. 463:3–20; Reider Tr. 2210:10–12.

This Court is unpersuaded by Lupin's argument of lack of nexus and finds that the secondary consideration of commercial success weighs against a finding of obviousness. Prezista could never have been launched as a commercial product—let alone become the success it has become—without a method for making bis-THF suitable for manufacture in industrial scale quantities. Leffler Tr. 2322:16–20. Bis–THF is a critical component of darunavir, accounting for 90 percent of the cost of manufacture. Wigerinck Tr. 279:20–25. Janssen uses the process of claim 1 of the '015 Patent to manufacture the bis-THF in Prezista. Reider Tr. 2025:8–15. This Court finds that there is a nexus between Prezista's commercial success and the claim 1 invention. See Tec Air, 192 F.3d at 1360–61 (finding a nexus between a product's commercial success and a patent on a method used to manufacture a component of the product); Akzo N.V. v. Int'l Trade Comm'n, 808 F.2d 1471, 1481 (Fed.Cir.1986) (finding commercial success where a product made by a patented method was commercially successful).

The Court finds no merit in Lupin's lack of nexus argument that doctors do not make prescription decisions based on the method of manufacturing bis-THF, Zingman Tr. 890:3–6, or its argument that Janssen does not promote the presence of bis-THF in darunavir in its marketing materials, Falcon Tr. 2494:15–17. Assuming that these contentions are true, they do not detract from the causal nexus between the method of making bis-THF and Prezista's commercial success. The success of darunavir, and Prezista, is attributable to multiple factors, with the manufacturing processes as critical steps in the chain of success. Stoffels Tr. 136:18–20 ("[W]e never would have had the drug if we didn't develop the drug and had invested over a billion dollars in this molecule.").

Because there is a nexus between claim 1 of the '015 Patent and the commercial success of Prezista, which was demonstrated at trial to be quite significant, this secondary consideration weighs against a finding of obviousness.

c. Praise and industry acceptance

Praise and industry acceptance provide additional evidence of nonobviousness. See Power–One, Inc. v. Artesyn Techs., Inc., 599 F.3d 1343, 1352 (Fed.Cir.2010) (praise); Pro–Mold & Tool Co. v. Great Lakes Plastics, Inc., 75 F.3d 1568, 1574 (Fed.Cir.1996) (industry acceptance). Prezista has received widespread praise: the guidelines for treating AIDS patients that are promulgated by the U.S. Department of Health and Human Services list Prezista as one of only two "preferred" protease-inhibitor based regimens for treating HIV/AIDS. PTX 592 (Guidelines for the Use of Antiretroviral Agents in HIV–1–Infected Adults and Adolescents) at F–4. This is the "highest possible recommendation" in the guidelines. Zingman Tr. 908:7–909:3. Prezista's acceptance among physicians is reflected by it being the leading protease inhibitor for treating AIDS. Leffler Tr. 2312:2–3. As with commercial success, none of this would have been possible without the method of making bis-THF that the '015 Patent describes and claims.

Lupin expresses the exact same lack of nexus arguments as it did with regard to commercial success. For the same reasons, the Court finds a causal nexus between claim 1 of the '015 Patent and the praise and industry acceptance of Prezista. The Court finds the showing Janssen made as to praise and industry acceptance persuasive such that these secondary factors also weigh against a finding of obviousness.

d. Copying

Finally, Janssen asserts that Lupin and Teva both copied the '015 invention and argues that such copying is another objective indicium of non-obviousness. Reider Tr. 2027:16–2028:20. Lupin and Teva both admitted using the method of claim 1 of the '015 Patent to manufacture bis-THF for their respective ANDA products. Summ. J. Op. at 54.

Lupin counters that in the context of Hatch–Waxman³ cases, the Federal Circuit has deemed evidence of copying by generic manufacturers to be "not probative of nonobviousness because a showing of bioequivalence is required for FDA approval." Bayer Healthcare Pharms., Inc. v. Watson Pharms., Inc., 713 F.3d 1369, 1377 (Fed.Cir.2013) ; see also Mitsubishi Chem. Corp. v. Barr Labs., Inc., 718 F.Supp.2d 382, 443–44 (S.D.N.Y.2010) ( "[C]opying bears little weight on its own as an objective indicator of non-obviousness in the context of generic drugs. The burdensome NDA [New Drug Application] procedures provide generic drug manufacturers with an incentive to copy an already approved drug, so that they can avail themselves of the less burdensome ANDA procedure of the Hatch–Waxman Act."). Lupin is correct, and as such the Court will not consider this secondary consideration of nonobviousness.

3 The Hatch–Waxman Act is officially the Drug Price Competition and Patent Term Restoration Act of 1984, Pub.L. No. 98–417, 98 Stat. 1585 (codified as amended in sections of 15, 21, 35, and 42 U.S.C.).

5. Conclusion: The '015 Patent was not obvious

This Court finds that claim 1 of the '015 Patent would not have been obvious to one of ordinary skill in the art under 35 U.S.C. § 103. Lupin has fallen well short of its mandate to provide evidence to support " ‘an abiding conviction’ " that any of its factual contentions are " ‘highly probable.’ " ActiveVideo, 694 F.3d at 1327 (quoting Colorado v. New Mexico, 467 U.S. at 316, 104 S.Ct. 2433 ). Rather, Lupin's case amounts only to a hindsight reconstruction of the invention.

In asserting that a retrosynthetic analysis would have "unambiguously" led a person of ordinary skill in the art to the starting point used in claim 1 of the '015 Patent, Dr. Ganem was improperly relying on hindsight. Dr. Ganem analogized his analysis to using a roadmap to find his way to his goal, Ganem Tr. 1009:19–1010:5–that is improper use of hindsight. When used to prove that an invention is obvious, working backwards presents a clear risk of hindsight. This Court is not aware of any case in which a court relied on retrosynthetic analysis as part of an obviousness analysis.

Secondary considerations provide additional evidence of nonobviousness. Here the objective considerations all confirm the nonobviousness of the '015 invention: failure by others, see DePuy Spine, Inc. v. Medtronic Sofamor Danek, Inc., 567 F.3d 1314, 1328 (Fed.Cir.2009) ; commercial success, Tec Air, 192 F.3d at 1360–61 ; praise and industry acceptance, Power–One, 599 F.3d at 1352.

ii. Enablement and Written Description under 35 U.S.C. § 112

Lupin argues that claim 1 of the '015 Patent is invalid under 35 U.S.C. § 112, ¶ 1 for not providing an enabling disclosure of the claimed invention. Lupin makes this argument in its Proposed Findings of Fact and Conclusions of Law, and also in its Memorandum in Support of its Rule 52(c) Motion for Judgment on Partial Findings of Invalidity Due to Lack of Enablement. ECF No. 843. This Court rejects Lupin's lack of enablement argument and finds that claim 1 of the '015 Patent is fully enabled, and it follows that this Court denies Lupin's Rule 52(c) Motion for Judgment on Partial Findings of Invalidity.

Lupin argues that, as a theoretical matter, there are eight stereoisomers of compound (7), but that only four exist (the specification affirmatively states that only four exist). Lupin contends that the specification does not enable the full scope of the claim—allegedly, a method for making eight stereoisomers of compound (7)—due to the fact that the claim does not disclose any stereochemistry.

As an initial matter, this Court finds that Lupin has made this argument in an untimely manner and therefore it must be rejected. This non-enablement argument is new and was not identified before trial nor mentioned in the Pre–Trial Order. See Janssen's Opp'n to Lupin's Rule 52(c) Motion at 2–6 (ECF No. 842). Under the express terms of the Pre–Trial Order and under Fed.R.Civ.P. 16(e), Lupin may not raise issues not disclosed in the Pre–Trial Order unless it can show "manifest injustice." Pre–Trial Order at 108.

Not only did Lupin fail to raise this issue before trial, it affirmatively embraced the accuracy and completeness of the '015 Patent's disclosure of bis-THF's stereoisomers. Lupin's expert, Dr. Ganem, speaking in reference to the European analog of the '015 Patent (which is identical in all relevant respects, and referred to as the "'853 publication"), stated in his expert report that "[i]t is readily apparent" that the patent "contemplates and discloses all possible stereoisomers of bis-THF." Deck of Ryan Mott, Ex. A at 49–50 (ECF No. 842). "In essence, the '853 publication represents a full disclosure of all thermodynamically stable bis-THF isomers." Id. at 55 [redacted]

Even if this argument were not untimely, it has no substance. Both the claim itself and the specification make clear that the compound of formula (7) is limited to the four actual stereoisomers of compound (7). The face of the claim shows a molecule with three chiral centers that is created by a "cyclization" reaction. The location of the chiral centers is such that the cyclization reaction can create only four stereoisomers, and this is apparent to one of skill in the art—even an undergraduate chemistry major—from the claim itself. Reider Tr. 2030:22–2031:6. Dr. Ganem did not disagree with this analysis. Ganem Tr. 1223:4–6.

If there were any doubt as to this issue, the specification removes it. The specification explains that eight stereoisomers of bis-THF "theoretically" are possible, but only four "are deemed to exist." DTX 7 ('015 Patent) at col. 9:8–14. In both words and drawings, the specification defines the compound of formula (7) to consist only of the four actual stereoisomers, which are denominated 7.1, 7.2, 7.3 and 7.4. Id. at col. 9:14–35. See AIA Eng'g Ltd. v. Magotteaux Int'l S/A, 657 F.3d 1264, 1276–79 (Fed.Cir.2011) (finding that patentee chose to apply a "special meaning to [a claim term]" where the ordinary meaning would have required a material that is "physically impossible to make"); SkinMedica, Inc. v. Histogen, Inc., 727 F.3d 1187, 1195–96 (Fed.Cir.2013) ("[I]f the specification reveals an intentional disclaimer, or disavowal, of claim scope by the inventor, the scope of the claim, as expressed in the specification, is regarded as dispositive." (citation omitted)); Wyeth v. Teva Pharms. USA, Inc., 03–cv–1293(WJM), 2005 WL 2175440, at *6 (D.N.J. Sept. 6, 2005) (adopting the narrower of two proposed constructions in reliance on statements in the specification that some embodiments were "impossible to achieve" and the principle that "[c]laims are not correctly construed to cover what was expressly disclaimed" (citation omitted)). When the compound of formula (7) is construed correctly, as covering the four stereoisomers of bis-THF that actually exist, there is no enablement issue and Lupin's enablement argument dissolves.⁴ Lupin also offers another invalidity theory, based on the "how to use" prong of the enablement inquiry. See Mem. in Support of Rule 52(c) Mot. for J. on Partial Findings of Invalidity Due to Lack of Enablement at 17–19; Joint Proposed Findings of Fact ¶¶ D421–38. Again, Lupin did not raise this theory at trial or in the Pre–Trial Order. In this enablement defense, Lupin does not dispute the adequacy of the '015 Patent's disclosure of the usefulness of compound 7.1 as a component of darunavir, but argues that the '015 Patent does not "identify a use for any of the other [stereo]isomers" of bis-THF. Mem. in Support of Rule 52(c) Mot. for J. on Partial Findings of Invalidity Due to Lack of Enablement at 18.

4 Lupin also argues that claim 1 of the '015 Patent does not meet the written description requirement of 35 U.S.C. § 112 because the claim covers isomers that are impossible to form, and one of ordinary skill in the art cannot accept that the inventors possessed a nonexistent invention. For the same reasons discussed, this argument is meritless. See, e.g., Wyeth, 2005 WL 2175440, at *6 ("[c]laims are not correctly construed to cover what was expressly disclaimed").

Again, having failed to raise this theory earlier, Lupin cannot raise it now. See Pfizer Inc. v. Teva Pharms. USA, Inc., 482 F.Supp.2d 390, 476 (D.N.J.2007) (barring defendants from relying on a new invalidity theory that was "never raised, or even alluded to, at trial or in the Pretrial Order"); see also Janssen's Opp'n to Lupin's Rule 52(c) Mot. at 2–6. But even if this Court were to consider this argument, it would find it meritless. The specification specifically describes bis-THF's use as a component of protease inhibitors:

[bis–THF] is an important pharmacological moiety present in the structure of retroviral protease inhibitors such as those described in Ghosh et al. in J. Med. Chem. 1996, 39(17), 3278–90, EP 0 715 618, WO 99/67417 and WO 99/76870. Said publications are hereby incorporated by reference.

DTX 7 ('015 Patent) at col. 1:18–24. This teaching is not limited to a specific stereoisomer of bis-THF.

Dr. Ghosh's 1996 Article (PTX 389)—"incorporated by reference" in the specification, id., and "effectively part of the ['015 Patent] as if it were explicitly contained therein," Cook Biotech, Inc. v. Acell, Inc., 460 F.3d 1365, 1376 (Fed.Cir.2006) (citation omitted)—provides potency data on a protease inhibitor with the 3(R),3a(S),6a(R) stereoisomer of bis-THF and another protease inhibitor with the 3(S),3a(R),6a(S) stereoisomer (which the '015 Patent refers to as compounds 7.1 and 7.3 respectively). See DTX 7 ('015 Patent) at col. 9:30–35. The Ghosh 1996 Article states that both had "antiviral potency" and "prevented the spread of HIV" in infected human cells. PTX 389 (Ghosh 1996 Article) at 3281. Furthermore, applying standard techniques, a person of ordinary skill could have determined without undue experimentation that protease inhibitors with all four stereoisomers of bis-THF have antiviral potency against the wild-type HIV virus. Such work is described in the Surleraux Article (PTX 114). Protease inhibitors with the 7.1, 7.2, 7.3 and 7.4 stereoisomers of bis-THF are referred to in the Surleraux Article as Compounds la, e, f and g respectively. Compare PTX 114 (Surleraux Article) with DTX 7 ('015 Patent) at col. 9:31–34. As described in the Surleraux Article, testing showed that protease inhibitors with all four stereoisomers of bis-THF have activity against the HIV virus.See PTX 114 (Surleraux Article) at 1817 (Table 3). There is no evidence that a person of ordinary skill reading the patent could not have obtained the same results at the time of the patent without undue experimentation.

Lupin's untimely "how to use" theory is unsupported by any evidence. Lupin did not elicit any testimony from any witness suggesting that a person of ordinary skill in the art would need undue experimentation to determine the utility of the four stereoisomers of bis-THF that can exist (or, for that matter, determining that the other four stereoisomers cannot exist). Lupin has a complete failure of proof.

The argument also fails on the law. Under black-letter law, a patent does not need to provide any test data showing the usefulness of any embodiment. Atlas Powder Co. v. E.I. du Pont De Nemours & Co., 750 F.2d 1569, 1576 (Fed.Cir.1984). "It is well settled that an invention may be patented before it is actually reduced to practice," i.e., before it is made, let alone tested. Alcon Research Ltd. v. Barr Labs., Inc., 745 F.3d 1180, 1189 (Fed.Cir.2014). " ‘[T]he burden is on one challenging validity to show by clear and convincing evidence that [a patent's] prophetic examples together with other parts of the specification are not enabling.’ " Id. (quoting Atlas Powder, 750 F.2d at 1577 ). Lupin has not met that burden.

Finally, Lupin's remaining argument—that bis-THF in isolation had "no known clinical use" (Mem. in Support of Rule 52(c) Mot. at 18)—is off target. The enablement issue is not whether bis-THF is useful by itself to prevent the spread of the AIDS virus or for other clinical purposes. As the specification explains, DTX 7 ('015 Patent) at col. 1:18–23, bis-THF is useful as a component of protease inhibitors. The specification provides an adequate disclosure of that utility. See DTX 7 at col. 1:18–24; DTX 389 (Ghosh 1996 Article) at 3281; see also Alcon, 745 F.3d at 1189 ; CFMT, Inc. v. Yieldup Int'l Corp., 349 F.3d 1333, 1337–39 (Fed.Cir.2003).

This Court finds that the '015 Patent is enabled and meets the written description requirement.

C. The '411 Patent

i. Obviousness

1. Level of Ordinary Skill in the Art

Janssen and Mylan offered separate definitions for a person having ordinary skill in the art. Dr. Reider, on behalf of Janssen, said that a person of ordinary skill in the art would have had a master's degree in chemistry and several years of laboratory experience. Reider Tr. 2003:4–7. Dr. Laird, on behalf of Mylan, opined that a person of ordinary skill in the art would have a Ph.D. in chemistry and several years' experience in related fields, such as organic chemistry or process chemistry. Laird Tr. 1383:11–18. The Court finds either of these definitions satisfactory and that regardless of which is selected, the analysis herein would be the same. Dr. Reider and Dr. Laird each stated that their opinions on validity did not depend on the definition of the level of ordinary skill in the art and would be the same under either side's definition. Reider Tr. 2003:17–20; Laird Tr. 1386:9–13. The Court adopts Dr. Reider's broader definition of a person of ordinary skill in the art which the trial evidence has shown to be pragmatically realistic.

2. Scope and Content of the Prior Art

Under Graham, the Court must define the scope and content of the prior art as of December 23, 2003, the priority date of the '411 Patent. Pre–Trial Order ¶ 45.

There is no dispute between the parties that there are three references that are part of the scope and content of the prior art with regard to the '411 Patent, and indeed each of these three references is asserted on the face of the '411 Patent. Laird Tr. 1457:2–18, 1489:21–25; DTX 13 at 2–3. These references are summarized below.

The '775 Patent (DTX 5): The '775 Patent, with an August 1994 priority date, discloses a class of "hydroxy ethyl amino sulfonamides," including darunavir. DTX 5 (title). In addition to describing the darunavir compound, the '775 Patent teaches a synthetic route that can be used for the synthesis of darunavir. Id. at cols. 64:47–65:9, 81:38–82:25, 82:53, 161:1–162:67; Laird Tr. 1467:9–13. The starting material used in the '775 Patent's scheme for synthesizing darunavir contains a "benzyloxycarbonyl," or "CBZ" protecting group. Reider Tr. 2039:17–21; Laird Tr. 1466:22–1467:8; PTX 948 (Laird diagram of '775 patent scheme). The ' 775 Patent calls the CBZ protecting group "a preferred amino protecting group." DTX 5 ('775 Patent) at col. 14:13–14.

The final step of the synthesis of darunavir in the '775 Patent, the "coupling" of bis-THF to the rest of the darunavir molecule, is carried out using DSC as the coupling agent. Laird Tr. 1398:3–13; DTX 5 ('775 Patent) at col. 82:52–54; Reider Tr. 2037:19–2038:1.

Janssen contends that another significant element of the '775 Patent is that it teaches simultaneous reduction and deprotection steps. Reider Tr. 2040:4–2041:14. Janssen asserts that combining the reduction of the nitro moiety and the deprotection of the CBZ-protected amino group into a single step was "kill[ing] two birds with one stone," id. at 2041:1–6, and would be seen as "highly efficient" by those of skill in the art. Id. at 2041:15–21. Mylan disputes this characterization of the reduction and deprotection steps in the ' 775 Patent. According to Mylan, one of ordinary skill in the art would not treat the reduction and deprotection steps of the '775 Patent as constituting a single chemical reaction. Rather, Mylan contends that two chemical reactions involving two different groups in two different structural locations take place irrespective of the identity of the protecting group and reagents used. Laird Tr. 1468:7–10, 1520:3–23. Mylan supports this position by stating that because the reduction reaction proceeds more quickly than the deprotection reaction on a molecular level, the simultaneous reduction/deprotection step of the '775 Patent is actually two steps. Laird Tr. 1519:19–1521:7. Mylan also claims that Janssen ignores that the reduction/deprotection steps can be performed as "one pot" reactions irrespective of the protecting group used, which minimizes the impact of the protecting group choice even if the number of steps was of import. Laird Tr. 1520:24–1521:7; Ebert Tr. 1308:3–17.

As of the priority date of the '411 Patent, a person of ordinary skill in the art would have viewed the synthesis disclosed in the '775 Patent as suitable for use on an industrial scale. Reider Tr. 2037:9–13; Laird Tr. 1464:19–25 ("I have said I thought it was scalable, yes."), 1472:17–1473:7 (the '775 scheme "would scale reasonably well").

The '889 (Ajinomoto) Publication (PTX 499): Ajinomoto is a major Japanese chemical company. The Ajinomoto Publication is discussed in detail in the specification of the '411 Patent. See DTX 13 ('411 Patent) at col. 2:10–30. It is a published patent application that discloses the same synthetic scheme as the '775 Patent. Reider Tr. 2038:7–14. Where the '775 Patent is directed to a class of compounds, the Ajinomoto Publication is addressed to industrial processes for synthesizing a single compound, the compound of formula (5), the penultimate compound before darunavir. Laird Tr. 1473:18–24; PTX 499 (title). The Ajinomoto Publication expressly teaches, by its reference to the Ghosh 1998 Article (discussed below), that this compound can be used to make HIV protease inhibitors including darunavir. PTX 499 at [0119].

The Ajinomoto Publication states that "an object of the present invention is to provide [ ] an industrially useful production method" for making what the ' 411 Patent calls compound (5). PTX 499 at [0008]. It says that Ajinomoto scientists "conducted intensive studies ... to develop ... [an] industrial production method," id. at [0009], and developed a method that can be used "industrially and efficiently," id. at [0141]. See also Reider Tr. 2037:9–15, 2043:12–20; Laird Tr. 1478:10–20, 1479:13–20.

The Ajinomoto Publication also discusses the benefits of "simultaneously performing" the "reduction" and "deprotection" reactions as a single process step. PTX 499 at [0090]; Laird Tr. 1477:3–7.

The Ghosh 1998 Article (DTX 1180): Like the Ajinomoto Publication and the ' 775 Patent, the Ghosh 1998 Article teaches a synthesis for making darunavir. Janssen asserts that like the '775 Patent and the Ajinomoto Publication, the Ghosh 1998 Article teaches a synthesis for making darunavir in which the reduction and deprotection reactions are combined in a single step. DTX 1180. Mylan again objects to this characterization of the reduction/deprotection steps, for the same reasons enumerated.

The Ghosh 1998 Article also teaches the use of DSC as a coupling agent to couple bis-THF with the sulfonamide at the final stage of the synthesis. Reider Tr. 2038:2–5; Laird Tr. 1463:4–1464:1. However, unlike Ajinomoto and the '775 Patent, Ghosh 1998 does not use a CBZ protected starting material, but instead uses an "azide" group. Laird Tr. 1462:14–18. Use of an azide group would "[set] some alarm bells ringing, because azides are well known to be potentially explosive, and some of them are shock-sensitive." Laird Tr. 1408:25–1409:17.

In addition to these three references, Mylan relies on two references that deal with amprenavir (an earlier generation HIV protease inhibitor, Laird Tr. 1393:5–9) called Kim and Tung, discussed below. This Court finds that neither of these references is analogous art that provides useful information for the synthesis of darunavir.

The Kim 2001 Article (DTX 604): The Kim 2001 Article discloses a 14–step, laboratory scale synthesis to make amprenavir. Reider Tr. 2078:1–6. Kim does not disclose darunavir or bis-THF. Its only alleged relevance in this case is its use of a boc protecting group. DTX 604 at 2351 (Scheme 2). A person of ordinary skill in the art would have had no reason to look to Kim in selecting a protecting group to use in the synthesis of darunavir.

The Tung '397 Patent (DTX 1001): The Tung patent discloses non-darunavir sulfonamides, including amprenavir. Reider Tr. 2076:17–20. The Tung patent lists 196 different compounds and processes for making them. DTX 1001 at col. 11–52. One of its 196 examples (Example 82) uses BNPC as a coupling agent. Reider Tr. 2076:9–16. The Tung patent does not discuss BNPC anywhere else in the specification. A person of ordinary skill in the art would have had no reason to look to Tung in selecting a coupling agent to use in the synthesis of darunavir.

3. Differences between the Asserted Claims and the Prior Art

The closest prior art methods for making darunavir have three key features: (1) simultaneous reduction and deprotection reactions, (2) use of a CBZ protecting group, and (3) the use of a DSC coupling agent. The method of claim 13 of the '411 Patent uses none of these features. Rather, claim 13 has separate reduction and deprotection steps, uses a boc protecting group, and uses a BNPC coupling agent. None of the prior-art methods for making darunavir —Ghosh 1998, the '775 Patent, or the '889 Ajinomoto Publication—teaches the use of any of these features, much less all of them in combination. Mylan's obviousness theory relies on combining the darunavir references with selected non-darunavir references (Kim and Tung). But Mylan failed to demonstrate, or even identify, any reason or motivation that would have led a person of ordinary skill in the art to search the non-darunavir art and choose these particular features to combine with the known methods of making darunavir. Mylan failed to prove that one of skill in the art would have any reason to change the known methods of making darunavir at all. Mylan has failed to meet its burden of establishing obviousness by clear and convincing evidence.

a. Separate reduction and deprotection steps

The inventors of the '411 Patent identified problems that the prior art had not recognized—e.g., exothermic reactions that could lead to explosions and catalyst poisoning—and solved these problems by separating the reduction and deprotection steps. Wigerinck Tr. 283:7–284:8; Reider Tr. 2047:1–10. The specification of the '411 Patent highlights this discovery in its discussion of the Ajinomoto process, which uses a simultaneous reduction/deprotection step with the CBZ protecting group. The specification describes how, in the Ajinomoto process,

This intermediate of interest is obtained ... [by forming the compound of formula (3) ], which is simultaneously reduced and deprotected to obtain the intermediate of interest.... It is observed that the simultaneous reduction ... and Cbz deprotection ... results in a highly exothermic reaction. Exothermic reactions, if possible, should be avoided.

DTX 13 ('411 Patent) at col. 2:10–30. The specification also identifies the catalyst poisoning problem with the Ajinomoto process. "A poisoned catalyst will inevitably result in the appearance of side-products thus decreasing product selectivity." Id. at col. 2:37–45. Mylan presented no evidence that suggests that a person of ordinary skill in the art would have seen a reason to abandon the simultaneous reduction/deprotection step disclosed in both the ' 775 Patent and the Ajinomoto Publication, and replace it with a process that separates these two steps, resulting in an extra step to the process.

At trial, Mylan argued that the simultaneous reduction/deprotection step of the '775 Patent and the Ajinomoto Publication is actually two steps. Dr. Laird asserted that the '775 Patent had distinct reduction and deprotection steps, for a total of "five steps." Laird Tr. 1404:13–21. He adhered to this view on cross-examination, but admitted that the Ajinomoto process had "a simultaneous reduction and deprotection step." Id. at 1475:9–10. It would be difficult to claim otherwise, as Ajinomoto states that it "simultaneously perform[s] elimination (deprotection) ... and reduction...." PTX 499 (Ajinomoto '889 Publication) at [0090].

This Court finds that the evidence adduced at trial shows that in the prior art '775 Patent and Ajinomoto references, the CBZ protecting group was removed (deprotected) under the "same conditions" by which the nitro moiety was reduced (specifically, a hydrogenation reaction), so that a single chemical step triggered both reactions, "kill[ing] two birds with one stone." Reider Tr. 2040:13–2041:6; see also Laird Tr. 1468:11–16 (agreeing that by using a hydrogenation reaction "you've chosen a reagent which happens to do those things at the same time, and that's an advantage for further development"). This hydrogenation reaction that was doing two things at the same time is what the prior art taught. In contrast, the '411 Patent calls for use of a protecting group (boc) which is not removed under reductive conditions, and, as a result, must be deprotected separately from the reduction step. Reider Tr. 2047:1–10, 2114:15–22. Thus this Court finds that the difference between the prior art processes and the '411 Patent is irrefutable. Two steps are required in the '411 Patent's process. That these two steps may be performed sequentially in a single vessel does not change the fact that they are separate steps. Unlike Ajinomoto and the '775 Patent, the method of claim 13 requires separate reduction and deprotection steps. Id. at 2119:7–16. The patent claims each as a separate "step"—reduction is "step (iii)" and deprotection is "step (iv)." DTX 13 ('411 Patent) at col. 23:7–50. Using two steps when the prior art taught one step runs counter to the conventional wisdom and increases time and cost. But it provides benefits in terms of safety and elimination of catalyst poisoning.

b. Boc protecting group

To repeat, the Ghosh 1998 Article uses an azide group to protect the second amino group in the synthesis of darunavir, while the '775 Patent and the Ajinomoto Publication both taught the use of a CBZ protecting group. Claim 13 of the '411 Patent does not use an azide group or the CBZ protecting group, but rather a boc protecting group. DTX 13 ('411 Patent) at cols. 23:52–24:50, 26:10–13. Before the '411 Patent, no one had ever used a boc protecting group in the synthesis of darunavir. Reider Tr. 2051:7–12. Unlike the dibenzyl and CBZ protecting groups, the boc protecting group does not come off during the reduction step that reduces the nitro moiety; this makes it possible to separate the reduction and deprotection steps of the darunavir synthesis when using a boc group. Wigerinck Tr. 285:18–286:2. Switching to the boc group contributed to solving the exotherm and catalyst poisoning problems that Tibotec experienced when attempting to reduce and deprotect in a single step. Id.

The change to a boc group also resulted in a yield improvement: 70 percent compared to the 47 percent yield produced by using the dibenzyl protecting group Tibotec had obtained from Searle. Wigerinck Tr. 290:25–291:11; PTX 427 (Tibotec Report on boc versus dibenzyl) at 2–3. The use of the boc protecting group also generated yields that were better than those reported in Ajinomoto, which used the CBZ protecting group in a synthesis suitable for industrial scale. Reider Tr. 2238:7–9; 2242:22–2243:3. From the data in the examples in the Ajinomoto Publication, one of ordinary skill in the art can calculate that the yield of its CBZ route "is worse" than the yield that Tibotec obtained using a boc group. Reider Tr. 2235:10–15.

Mylan presented no evidence that one of skill in the art would have had any reason or motivation to use boc, and no reason to expect the yield increases that resulted from that choice. As shown, while those of skill in the art would have been motivated to—and did—replace the azide group of the Ghosh 1998 Article with the CBZ group of the '775 Patent and the Ajinomoto Publication, Laird Tr. 1409:19–1410:2, 1462:21–23, nothing in the art suggested any reason to replace the CBZ group of the '775 Patent and Ajinomoto. On the contrary, Dr. Laird conceded that these processes would have been expected to "scale reasonably well." Laird Tr. 1472:17–1473:3.

Mylan has argued that the '775 Patent's general identification of boc as an alternative protecting group, without any reference to darunavir, renders it an obvious alternative. DTX 5 ('775 Patent) at col. 14:13. But one needs a reason or motivation to make a switch, see infra Section II(C)(i)(5), and Mylan does not present one. The '775 Patent expressly identifies CBZ, not boc, as a "preferred" choice for the synthesis of darunavir. DTX 5 ('775 Patent) at col. 14:13–14, 81:38–82:25.

Mylan also points to the Kim 2001 Article, which teaches the use of boc in the synthesis of amprenavir—but Kim does not provide any reason to use boc in the synthesis of darunavir. First, the synthetic route for amprenavir disclosed in Kim is completely different from the known routes for darunavir, and does not involve making the compound of formula (5) as in the '411 Patent. The Kim 2001 Article does not discuss darunavir at all. Moreover, Kim identifies no particular benefits of the boc group. Dr. Reider credibly testified that if Kim were considered at all, it would have been seen as unhelpful for developing an industrial scale darunavir synthesis because it teaches an unattractively lengthy synthetic scheme (14 steps) for a different molecule, and uses toxic tin chloride. Reider Tr. 2078:9–2079:11. This Court finds that all that Kim teaches is that a boc protecting group has been used in the past to make a different compound, but it does not provide any reason to combine boc with the darunavir art.

c. BNPC coupling agent

All of the prior art references that discuss darunavir —the '775 Patent, the Ajinomoto Publication, and the Ghosh 1998 Article—either expressly taught a DSC coupling agent or incorporate such a teaching by reference. DTX 5 ('775 Patent) at col. 82:53; Laird Tr. 1398:3–13, 1476:8–11, 1463:9–12; PTX 499 (Ajinomoto Publication) at [0119]; DTX 1180 (Ghosh 1998 Article) at PREZ00302761. Those references did not teach any reason to seek an alternative to DSC. Reider Tr. 2055:8–10. None taught or suggested the use of BNPC, the coupling agent in claim 13.

Tibotec found that the benefits of using BNPC were substantial—specifically, a significant increase in yield in the valuable last step and greater purity of the end product. Tibotec discovered that BNPC generated yields up to 80 percent for the final coupling step, compared to 71 percent for DSC. PTX 433 (Chem–Pharm Team Meeting Notes) at PREZ05041021. BNPC also allowed for a cleaner impurity profile and provided a more robust process. Id.; see Laird Tr. 1505:8–25, 1506:19–24. An internal memorandum confirmed that "[t]his coupling process [BNPC] has major advantages over the use of DSC as [a] coupling reagent: [1] it allows to increase the yield significantly, which has a major impact on the COGS [cost of goods sold;] [and 2] the quality of the DS [drug substance] is increased significantly...." PTX 433 (Chem–Pharm Team Meeting Notes) at PREZ05041021. These superior qualities of BNPC are reflected in the '411 Patent. Examples 7 and 8 show that BNPC has a higher purity than DSC and a higher yield in experiments using the same reagents. DTX 13 ('411 Patent) at col. 21:46–22:40; Reider Tr. 2055:19–2059:25.

Mylan's arguments as to why the use of BNPC is obvious are contrary to the documentary record. First, Mylan notes that Tibotec did not ultimately use BNPC in its commercial process. Tibotec counters that by the time it discovered BNPC, it was "too late to introduce the change [from DSC] in time" for it to be used in Tibotec's application for regulatory approval of darunavir. Laird Tr. 1505:2–25, 1506:21–25; Wigerinck Tr. 296:18–297:1; Reider Tr. 2060:6–16. Mylan challenges this by stating that, to this day, Janssen has not amended its process to use BNPC as a coupling agent, which calls its claims as to its superiority into question. But that Janssen has not chosen to amend its process to use BNPC is not dispositive of anything, and this Court cannot conjecture as to why it has made that choice. Tibotec/Janssen's use of DSC does not erase the advantages of BNPC that were adduced at trial, advantages which Mylan itself has recognized.

Second, Mylan disputes that the examples in the '411 Patent show the superiority of BNPC. It argues that Example 9 showed the possibility of an 81 percent yield with DSC and claimed that was comparable to the 81 percent yield obtained with BNPC in Example 8. Laird Tr. 1440:5–1441:14. But Dr. Reider credibly explained that that comparison was inapt. First, the BNPC in Example 8 was dissolved in triethylamine, while the DSC in Example 9 was dissolved in pyridine. Based on the data he examined, Dr. Reider stated that he would expect an increase in yield with BNPC if it were dissolved in pyridine. And beyond the yields, the DSC in Example 9 was only 95 percent pure, while the BNPC in Example 8 was 100 percent pure, meaning that BNPC resulted in a purer end product in any event. Reider Tr. 2059:4–25. Examples 7 and 8 were the best comparisons and they are completely consistent with Tibotec's internal documents showing the same results. Id. at 2244:15–2245:2.

The sole prior art reference Mylan relies on for the use of BNPC is the Tung patent. The Tung patent is not about darunavir. It discloses amprenavir and other non-darunavir sulfonamides whose synthesis is not analogous to the synthesis of darunavir. Only one of the 196 different compounds in Tung uses BNPC as a coupling agent, and there is no suggestion of any benefit associated with its use. Laird Tr. 1510:6–1512:2. There would have been no reason for a person of ordinary skill in the art to look to Tung for a coupling agent to use in the synthesis of darunavir. Mylan used hindsight to reconstruct the invention by looking to the prior art for BNPC, and as a result it cites a single mention of BNPC in a non-analogous patent. Reider Tr. 2077:13–15.

Tibotec's experience confirms what an analysis of the prior art shows—that replacing DSC with BNPC was not obvious. Tibotec made a concerted effort to find an alternative to DSC. The contractor it retained to study the issue concluded that "commercially available alternative coupling agents with promising reactivity were not uncovered." PTX 434 (Carbogen Process Research Report) at PREZ5073327. It was not until two years later that Tibotec, on its own, discovered BNPC and its advantages.

4. Secondary Considerations

a. Unexpected results

Janssen asserts that the prior art provided no reason for one of skill in the art to alter the known methods of making darunavir and arrive at claim 13 of the '411 Patent, and that the prior art provided no reason to expect that there would be any advantage in doing so. However, Janssen claims, it turned out that using the boc protecting group and the BNPC coupling agent had advantages over prior art methods that were entirely unexpected. Those unexpected results were that the boc protecting group and the BNPC coupling agent provided enhanced yield and purity, while reducing costs. Janssen contends that these unexpected results are objective indicia of nonobviousness.

Mylan counters by stating that the percent yield increases discussed by Janssen are not the type of results that were unexpected. Mylan cites to Federal Circuit precedent that "[u]nexpected results that are probative of nonobviousness are those that are different in kind and not merely in degree from the results of the prior art." Galderma Labs., L.P. v. Tolmar, Inc., 737 F.3d 731, 739 (Fed.Cir.2013). "Results which differ by percentages are differences in degree rather than kind, where the modification of the percentage is within the capabilities of one skilled in the art at the time." Id. Mylan argues that the improved yields Janssen cites are based on improper comparisons, and that even if such improvements were indeed real, they are not the type of result that is different in kind rather than degree.

First, with regard to protecting group yields, Mylan states that the '411 Patent does not provide any head-to-head comparative yield data of the product of the boc chemical reactions with those that used the CBZ protecting group. See DTX 13 ('411 Patent). Mylan contends that yield improvements failed to materialize when using the Ajinomoto Publication as a comparator. PTX 499 (Ajinomoto Publication). When considering the upper and lower yield ranges reported in the '411 Patent for the synthesis of the entire "right hand" side of the molecule, multiplying out the lower yield ranges using boc produced a 65.25 percent overall yield. Reider Tr. 2217:25–2218:6. Using the upper ranges produced a 77.35 percent overall yield. Id. at 2218:7–14. The reported overall yield to produce the same structure via the CBZ protecting group by way of the Ajinomoto Publication was 61.9 percent overall yield on the low end and approximately 80 percent on the high end. Id. at 2216:2–7, 2216:25–2217:4. Mylan asserts that these ranges squarely overlap and there is no evidence of any meaningful difference between the yields achievable with CBZ versus boc in a way which would indicate the boc results were unexpectedly superior.

With regard to coupling agents, Mylan asserts that Dr. Wigerinck's testimony on direct examination that the use of BNPC as the coupling agent led to better yields than DSC was based on his recollection, and not any documentary evidence presented at trial. Wigerinck Tr. 417:16–418:12. Mylan also points out that Dr. Wigerinck was not personally involved with the relevant research work, and that work was conducted by the Schaffhausen group within Johnson & Johnson, including by Hartmut Zinser and Birgit Ebert. Id. at 418:19–419:18. Mylan states that Dr. Wigerinck pointed to one test document from Carbogen, a third-party laboratory Tibotec contracted with, where the coupling agents were varied. In these experiments, the coupling reactions were not shown to outperform DSC. PTX 434 (Carbogen report); Wigerinck Tr. 410:20–411:17.

Mylan also attacks Dr. Wigerinck's claim that it was significant that the yield reported for Example 7 of the '411 Patent (using DSC as the coupling agent) was 71 percent while the yield for Example 8 of the '411 Patent (using BNPC as the coupling agent) was reported to be 81 percent. DTX 13 ('411 Patent) at col. 22:8–9; 39–40; Wigerinck Tr. 543:9–544:5. Mylan states that this is a false comparison because the comparison should be limited to the first part of the process, which relates to production of crude darunavir (i.e., before the subsequent recrystallization step which also affects yield). Laird Tr. 1440:5–1441:2. Mylan claims that the yield for the coupling procedure is the same for Examples 7 and 8: 43.5 grams. DTX 13 ('411 Patent) at col. 22:1, 32; Wigerinck Tr. 415:10–14, 416:2–14, 416:24–417:3; Laird Tr. 1440:5–1441:2. The same amount of starting material was used in both Examples. DTX 13 ('411 Patent) at col. 21:52; col. 22:17; Wigerinck Tr. 416:24–417:3.

Finally, Mylan argues that Janssen's arguments that BNPC is superior to DSC are undermined by the circumstance that Janssen itself does not use BNPC for its commercial process. Janssen claims that it was too late for them to adopt BNPC based on the timing of regulatory filings. Wigerinck Tr. 296:17–297:1, 420:10–14. Mylan disputes this position arguing that the NDA for Prezista was submitted in December 2005, over two-and-a-half years after the benefits of BNPC were discussed at a Chem–Pharm meeting. Wigerinck Tr. 424:6–425:20; PTX 433 (Apr. 11, 2003 Chem–Pharm Minutes) at PREZ05041021. Mylan further says that even if that initial window was missed, the FDA allows process changes to be made after submission of regulatory filings, Laird Tr. 1439:16–21, and that to this day, Janssen has never attempted to pursue BNPC as a coupling reagent for its product despite Dr. Wigerinck's claim that doing so would have a "major impact on the COGS [cost of goods sold]." Wigerinck Tr. 297:20–298:3. Mylan asserts that Janssen subsequently made other changes to the proposed commercial process for making the darunavir active pharmaceutical ingredient in Prezista that involved changing a solvent and using a pyridine reagent. Laird Tr. 1439:3–15. As a result, Mylan argues that there was no reason why Janssen could not have implemented the change from DSC to BNPC if it was genuinely advantageous to do so, and that because Janssen has not done so, Dr. Wigerinck's assertion that the yields attained would have constituted a significant improvement is not credible.

The Court finds that the '411 invention did produce unexpected results and that this secondary consideration weighs against a finding of obviousness. As to the differences in yields between using a boc protecting group rather than CBZ, this Court agrees with Janssen that the use of a boc protecting group generated yields superior to those reported in Ajinomoto, which used the CBZ protecting group in a synthesis suitable for industrial scale. Reider Tr. 2238:7–9, 2242:22–2243:3. From the data of the Ajinomoto Publication's examples, one of ordinary skill in the art can calculate that the yield of its CBZ route "is worse" than the yield that Tibotec obtained using a boc group. Reider Tr. 2235:10–15.

With regard to coupling agents, this Court also agrees with Janssen that the benefits of using BNPC over DSC were substantial and unexpected. The superior qualities of BNPC are reflected in the '411 Patent. Examples 7 and 8 show that BNPC has a higher purity than DSC and a higher yield in controlled experiments using the same reagents (that is, everything in the experiment, including the crystallization steps, was exactly the same, except one used DSC and one used BNPC). DTX 13 ('411 Patent) at col. 21:46–22:40; Reider Tr. 2055:19–2059:25. These examples show that BNPC is a superior coupling agent to DSC. Reider Tr. 2244:15–2245:2. As to Mylan's argument that Janssen's position that BNPC is superior to DSC is undermined because Janssen does not use BNPC for its commercial process, this Court is not convinced. This Court does not conjecture as to why Janssen has chosen not to use BNPC in its synthesis up to this point, but that it does not is of no moment. Evidence presented at trial clearly demonstrates the unexpected result that BNPC is superior to DSC. That is what matters for the inquiry under this secondary consideration.

As to Mylan's argument that the unexpected results demonstrated by Janssen relate to degree rather than kind, again this Court is not convinced. The case law Mylan cites refers to results that are changes of degree instead of kind because they are achieved through routine optimization, including varying elements like temperature and concentration available to a person of skill in the art. See, e.g., Galderma, 737 F.3d at 739. Here, the Court is not dealing with those types of variations, but rather with variations of different compounds, and Janssen has shown that the modification of the yield percentages were not "within the capabilities of one skilled in the art at the time." Id. The trial evidence here does not indicate that skilled artisans were capable of achieving greater yield by adjusting and controlling factors that were known to affect the yield percentages.

b. Copying

Janssen also argues that the fact that Mylan practices claim 13 of the '411 Patent is a secondary consideration that weighs against a finding of obviousness.

Mylan responds, as Lupin did with regard to the '015 Patent, that in the context of HatchWaxman cases, the Federal Circuit has deemed evidence of copying by generic manufacturers to be "not probative of nonobviousness because a showing of bioequivalence is required for FDA approval." Bayer, 713 F.3d at 1377. Again, Mylan is correct. The Court does not consider this secondary consideration of nonobviousness.

5. Conclusion: The '411 Patent was not obvious

Under the governing legal standards for obviousness, Mylan has failed to establish by clear and convincing evidence that the method of claim 13 would have been obvious to a person of ordinary skill in the art. Under Mylan's obviousness theory, a person of ordinary skill in the art would have rejected an acceptable manufacturing process for making darunavir taught in the prior art, despite the absence of any reason for doing so, and would then have formulated a longer and more expensive process that combined aspects of the prior art methods with isolated features chosen from non-analogous, non-darunavir references. Without the benefit of hindsight, there is no reason why a person of ordinary skill in the art would have done that.

As an initial matter, Mylan has failed to identify any reason or motivation that would have led a person of ordinary skill in the art to depart from the known methods for making darunavir as taught, especially by the '775 Patent and the Ajinomoto Publication. As Mylan's expert even conceded, the methods taught in those references were "pretty good." Laird Tr. 1472:17–1473:3. The prior art provided no motivation for a person of ordinary skill in the art to change them.

Specifically, Mylan has failed to give a reason why a person of ordinary skill would have selected the elements that Mylan found in two non-darunavir references (the Kim 2001 Article and the Tung patent) and combined those elements with known methods for making darunavir. The only reason to do so would be based on hindsight. Mylan did not even seek to prove a reason or motivation to combine references, but rather argued that it did "not need to prove an independent motivation" under KSR Int'l Co. v. Teleflex, Inc., 550 U.S. 398, 127 S.Ct. 1727, 167 L.Ed.2d 705 (2007), because evidence of motivation is not required when an obviousness case involves a combination of references known to people of skill in the art. Tr. 2555:8–2557:1 (closing argument). Such a reading of KSR is incorrect.

The Federal Circuit has repeatedly rejected arguments that KSR eliminated the need to prove a reason or motivation to combine references. As the Federal Circuit has emphasized, "[a] reason for combining disparate prior art references is a critical component of an obviousness analysis." InTouch Techs., Inc. v. VGO Commc'ns, Inc., 751 F.3d 1327, 1351 (Fed.Cir.2014). See also Otsuka Pharm. Co., Ltd. v. Sandoz, Inc., 678 F.3d 1280, 1292 (Fed.Cir.2012) ("The challenger of the patent must show by clear and convincing evidence that a skilled artisan would have been motivated to combine the teachings of the prior art references to achieve the claimed invention, and that the skilled artisan would have had a reasonable expectation of success in doing so." (citation omitted)); Procter & Gamble Co. v. Teva Pharms. USA, Inc., 566 F.3d 989, 994 (Fed.Cir.2009) (same). The Federal Circuit has clearly written that the KSR Court "acknowledged the importance of identifying ‘a reason that would have prompted a person of ordinary skill in the relevant field to combine the elements in the way the claimed new invention does' in an obviousness determination." Takeda Chem. Indust., Ltd. v. Alphapharm Pty., Ltd., 492 F.3d 1350, 1356–57 (Fed.Cir.2007) (quoting KSR, 550 U.S. at 418, 127 S.Ct. 1727 ). Under KSR, the reason to combine need not be explicit in the prior art, but there must be a reason. KSR, 550 U.S. at 418–19, 127 S.Ct. 1727. In fact, since KSR, the Federal Circuit has rejected obviousness defenses for failure to provide a reason or motivation to combine references. See, e.g., InTouch Techs., 751 F.3d at 1349 (reversing obviousness determination where expert "did not explain what reason or motivation one of ordinary skill in the art at the time of the invention would have had to place these pieces together"); ActiveVideo, 694 F.3d at 1328 (affirming finding of nonobviousness where defendant's expert "fail[ed] to explain why a person of ordinary skill in the art would have combined elements from specific references").

Mylan also had the burden of showing by clear and convincing evidence that a person of ordinary skill in the art would have had a "reasonable expectation of success" in making that combination of references. See Otsuka, 678 F.3d at 1292 ; Cyclobenzaprine, 676 F.3d at 1069. Mylan failed to meet that burden. The inventors of the '411 Patent solved problems that the prior art did not even recognize, let alone solve. The inventors' identification of those problems and their novel solutions were nonobvious contributions. As the Federal Circuit recently observed, "an invention can often be the recognition of a problem itself." Leo Pharm. Prods., 726 F.3d at 1353. The solutions chosen by the inventors went against conventional wisdom, adding both time and expense to the process by separating one step into two.

Mylan's theory is based on hindsight, and its argument that a person of ordinary skill would have selected boc from the Kim 2001 Article and BNPC from the Tung patent, and then combined those teachings with aspects of the darunavir art, is "hindsight reconstruction" of the invention that the Federal Circuit condemns. In re NTP, Inc., 654 F.3d 1279, 1299 (Fed.Cir.2011). Evidence of secondary considerations helps courts "avert the trap of hindsight." Leo Pharm. Prods., 726 F.3d at 1358 (citation omitted). The relevant secondary considerations here, including unexpected results of improved yield and purity, weigh against a finding of obviousness.

Considering all of the Graham factors, this Court finds that Mylan has failed to establish its obviousness defense by clear and convincing evidence.

ii. Enablement and Written Description under 35 U.S.C. § 112

Mylan also raises enablement and written description defenses under 35 U.S.C. § 112, ¶ 1. Claim 1 of the '411 Patent (from which claim 13 depends) recites a process for making darunavir (identified by its chemical structure) "or an addition salt, thereof." Mylan's theory is that one of ordinary skill in the art would need undue experimentation to make a salt of darunavir and, for that reason, the patent was not enabled. Tr. 2560:7–11 (closing argument).

Mylan did not meet its burden of showing that practicing claim 13 would require undue experimentation. Both Dr. Reider and Dr. Laird agreed that a person of ordinary skill in the art knew how to make salts and that the '411 Patent gives many examples. See, e.g., Reider Tr. 2080:11–16 ("Q. And did one of ordinary skill in the art know how to make a salt of darunavir from the disclosure? A. Yes. An addition salt, frankly, [is] just that: You add two things together to make the salt. You take your base and an acid, you mix them together. That's how you make a salt. Q. Are there multiple salts disclosed in the patent? A. There are."); Laird Tr. 1513:8–14 ("Q. Those with skill in the art know how to make salts; correct? A. Yes.... Q. The specification gives many examples; correct? A. It does, yes."); see also DTX 13 ('411 Patent) at cols. 16:64–17:26 (disclosing dozens of salts). As part of his work in this case, Dr. Reider reviewed experiments conducted by Tibotec creating seven darunavir salts, including hydrobromide, hydrochloride, the sulfate, the methane sulfonate, and the ethane sulfonate. Reider Tr. 2085:3–8.

Mylan's main argument was that the patent did not enable making pharmaceutically useful addition salts of darunavir. Laird Tr. 1444:14–19. This argument has no substance. First, Dr. Reider credibly testified without contradiction that although the salt forms of darunavir fabricated by Tibotec were admittedly not as good as the ethanolate form, they could still be put into a capsule and given to a patient if there was nothing else available. This is what Merck did when it could not formulate its ethanolate salt of Crixivan into a tablet, and it provided benefits to AIDS patients. Reider Tr. 2080:19–2081:5. For that reason, the '411 Patent does enable the manufacture of pharmaceutically useful salts. But more importantly, Mylan's premise is wrong. Claim 13 does not require a pharmaceutically useful salt, only a salt. As the patent teaches, there are many uses for darunavir salts that do not require that the salt be pharmaceutically useful. The patent states that "salts having a pharmaceutically unacceptable counter—ion may also find use, for example, in the preparation or purification of a pharmaceutically acceptable compound of the present invention." DTX 13 ('411 Patent) at col. 16:58–62. Non-pharmaceutically acceptable salts, used in a manufacturing process, can provide a "tremendous improvement in purity, chirality, and control." Reider Tr. 2081:9–2082:9. The '411 Patent therefore teaches that "[a]ll salts, whether pharmaceutically acceptable or not, are included within the ambit of the present invention." DTX 13 ('411 Patent) at col. 16:62–63; Reider Tr. 2081:6–8.

The undisputed evidence shows that those of skill in the art were able to make salts of darunavir based on the disclosures in the '411 Patent, and Mylan failed to demonstrate by clear and convincing evidence that a person of ordinary skill in the art, reading the '411 Patent, would not be able to make an addition salt of darunavir by known methods.

Finally, Mylan argues that the '411 Patent fails the written description requirement. It argues that for the same reasons regarding enablement, the inventors were not in possession of processes for the preparation of addition salts of darunavir. Joint Proposed Findings of Fact ¶ D730. And for the same reasons declared by the Court regarding enablement, this argument is rejected. Trial evidence shows that the inventors of the '411 Patentwere in possession of processes for the preparation of addition salts. The written description requirement is met. Second, Mylan asserted only in closing arguments that the BNPC element of claim 13 is not supported by the patent's written description because "[t]here's no direction in the patent to single out BNPC as a coupling reagent." Tr. 2559:3–4 (closing argument). That assertion is wrong. BNPC is expressly identified in the specification (col. 15:48), is used in experimental Example 8 and is claimed in claim 13. That establishes that Tibotec was in possession of the invention that is claimed. Mylan did not present any trial evidence in support of this challenge and, as a result, failed to demonstrate it.

The Court finds that the '411 Patent is valid and rejects Mylan's enablement and written description arguments.

D. The '645 Patent

i. Obviousness

1. Level of Ordinary Skill in the Art

Janssen and Lupin offered separate definitions for a person having ordinary skill in the art. Dr. Allan Myerson, on behalf of Janssen, said that a person of ordinary skill in the art would have had a bachelor's degree in chemistry or related disciplines with three or more years of experience, or a master's degree in the same field with two or more years of experience, or a Ph.D. in the same field with no additional experience. Myerson Tr. 1780:11–18. Dr. Zaworotko, on behalf of Lupin, opined that a person of ordinary skill in the art would have a Ph.D. in chemistry or a related discipline and several years of experience, or a bachelor's or master's degree and specific training in the area of crystallization. Zaworotko Tr. 1592:3–20. The Court finds either of these definitions satisfactory and that regardless of which is chosen, the analysis herein would be the same. Dr. Myerson and Dr. Zaworotko each stated that their opinions on validity did not depend on the definition of the level of ordinary skill in the art and would be the same under either side's definition. Myerson Tr. 1780:19–23; Zaworotko Tr. 1592:21–1593:3. The Court adopts Dr. Myerson's broader definition of a person of ordinary skill in the art which the trial evidence has shown to be pragmatically realistic.

2. Scope and Content of the Prior Art

Janssen asserts that the prior art for the '645 Patent includes two references, summarized below, disclosing the darunavir compound:

The Searle '775 Patent (PTX 9): The Searle '775 Patent is listed on the '645 Patent's cover page as a reference that the Patent Examiner considered, and the specification incorporates it by reference. PTX 1 ('645 Patent) at col. 1:49–55. The '775 Patent suggests various formulations for the compounds it discloses. See PTX 9 at col. 216:23–217:53. Those possible formulations include "salts," id. at col. 216:23–26, "esters," id. at col. 218:12–22, "prodrugs," id. at 217:54–65, "injectable aqueous or oleaginous suspensions" and other "liquid dosage forms," id. at col. 217:16–19, 217:46–50. Salts, esters and prodrugs are also claimed in the '775 Patent. Zaworotko Tr. 1685:21–1686:1; PTX 9 at col. 218:12–22. Ethanolates and solvates are not among the numerous formulations for darunavir that the ' 775 Patent teaches and suggests. Myerson Tr. 1902:4–13.

The Ghosh 1998 Article (DTX 1180): The Ghosh 1998 Article is listed on the ' 645 Patent's cover page as a reference that the Patent Examiner considered, and the specification incorporates it by reference. PTX 1 ('645 Patent) at col. 1:49–55. The Ghosh 1998 Article does not describe any formulation for any of the compounds it discusses. See Myerson Tr. 1902:4–13.To support its obviousness argument, Lupin also relies on various references that relate generally to: solid-state screens; the use of ethanol as a solvent; and crystalline forms of salts or intermediates of other protease inhibitors. See generally Joint Proposed Findings of Fact ¶¶ D836–80. None of those references mentions darunavir, or describes or suggests a solvate of darunavir, and none of them provides any guidance on the conditions needed to form an ethanolate solvate of darunavir.

3. Differences between the Asserted Claims and the Prior Art

Nothing in the prior art taught or suggested that darunavir could be made in an ethanolate form or that an ethanolate form of darunavir would be desirable. To the extent that the prior art discussed darunavir, it suggested other formulations, like salts, esters, and prodrugs, and did not suggest a solvate or ethanolate form. In addition, solvates are relatively rarely used for pharmaceutical products and there is an aversion to using them when it can be avoided. Furthermore, solvate formation is very unpredictable. The art did not provide a basis for a reasonable expectation that an ethanolate of darunavir could be created, and if it could be created, there was no teaching in the art of the combination of conditions that would be needed to create it. And if it were possible to find the right combinations to create a solvate of darunavir, the prior art offered no reason to expect that it would exist in a 1:1 ratio of darunavir to ethanol or have properties that would make it work as a pharmaceutical composition.

At trial, Lupin's expert Dr. Zaworotko presented a "five-step pathway" to obviousness, which purportedly leads to the conclusion that claim 4 of the ' 645 Patent would have been obvious to a person of ordinary skill. The Court finds that such a path is ended before started.

The first step in Dr. Zaworotko's "Obviousness Pathway" was that the "[d]arunavir molecule [was] known before 2002." Zaworotko Tr. 1594:2–9. The Searle '775 Patent and the Ghosh 1998 Article both describe darunavir, and the Patent Examiner considered both references. The Ghosh 1998 Article does not suggest any formulation of darunavir, and the '775 Patent suggests numerous formulations (e.g. salts, esters and prodrugs), but does not mention solvates or an ethanolate. Although darunavir was known in the art, such does not establish this first step. Dr. Zaworotko explained, "the first step in [his] pathway" was an assertion that "darunavir was a known molecule and ... [was] a compound of interest" by 2002, meaning that a person of ordinary skill in the art would have been interested in making a formulation of darunavir out of the millions of possible compounds known in 2002. Id. at 1595:4–9. Demonstrating that darunavir was a compound of interest in 2002 was the essential part of Lupin's burden in alleging obviousness. Otsuka Pharm. Co. v. Sandoz, Inc., 678 F.3d 1280, 1291 (Fed.Cir.2012) ( "[W]hether a new chemical compound would have been prima facie obvious over particular prior art compounds ordinarily follows a two-part inquiry. First, the court determines whether a chemist of ordinary skill would have selected the asserted prior art compounds as lead compounds, or starting points, for further development efforts."); Unigene Labs., Inc. v. Apotex, Inc., 655 F.3d 1352, 1361–62 (Fed.Cir.2011) ; Eisai Co. v. Dr. Reddy's Labs., Ltd., 533 F.3d 1353, 1359 (Fed.Cir.2008) ("[P]ost KSR, a prima facie case of obviousness for a chemical compound still, in general, begins with the reasoned identification of a lead compound"). As Dr. Zaworotko agreed, "[i]f the person of [ordinary] skill wouldn't have had any motivation to work on darunavir, then they certainly wouldn't have found any new crystal forms of it...." Zaworotko Tr. 1682:5–10.

Lupin failed to show that darunavir would have been a compound of interest in 2002. Dr. Zaworotko asserted that darunavir would have been a "compound of interest" because it "was in clinical trials as of late 2001." Zaworotko Tr. 1595:18–23; see Stoffels Tr. 98:7–17. Earlier, this Court had relied on this proposition to deny summary judgment. See Summ. J. Op. at 16. But that argument evidentially collapsed at trial. Although darunavir was in clinical trials in 2002, that fact was highly confidential and could not have provided a motivation to anyone outside of Tibotec to select darunavir for formulation. As Dr. Stoffels explained, the clinical trials on darunavir were kept so secret that Tibotec provided the initial results to Johnson & Johnson in 2002 only under a confidentiality agreement. Stoffels Tr. 98:15–25. Dr. Reider, the former Vice President of Process Development at Merck, confirmed that secrecy about drugs under development is the norm in the industry because the pharmaceutical industry is highly competitive and the pharmaceutical companies "do not necessarily share all of their ... successes and their secrets" with their competitors. Reider Tr. 1993:14–24. Because clinical trials are not always successful, pharmaceutical companies also keep them secret because they do not want to raise the expectations of the marketplace and the patient population prematurely. Id. at 1993:20–1994:5.

Lupin needed to point to something in the public record or in the prior art reporting that darunavir was in clinical trials as of the '645 Patent's priority date of May 16, 2002 in order to show that darunavir would have been a compound of interest. Lupin failed—Tibotec's development of darunavir and its clinical trials were secret in 2002. The only mention in the prior art of clinical trials of darunavir was a single sentence in a lengthy article by Dr. Yoshimura. The article is about TMC 126, the compound preferred by Dr. Ghosh and Dr. Erickson, which is also known as UIC 94–003. The last sentence of the article reads: "A potent analog of UIC 94–003, TMC 114, with a similar resistance profile, is currently undergoing clinical trials...." DTX 621 (Yoshimura 2001 Article) at 0621. But this sentence provides no relevant information to one of skill in the art. There is no clue in the article, or anywhere else in the prior art, what TMC 114 is. The "TMC" designations were Tibotec's "internal code name[s]" for compounds, Stoffels Tr. 142:10–12, and TMC 114 was Tibotec's "internal identifier" for darunavir. Id. at 128:11–15; see Wigerinck Tr. 359:20–25.Darunavir is not even specifically identified. The designation does not permit any third party to know the compound's chemical structure and so could not possibly provide a motivation for anyone to attempt to crystallize TMC 114. Absent public notice that darunavir was in clinical trials, Lupin's "compound of interest" theory vanishes—and with it Lupin's obviousness challenge. This Court denied summary judgment on this issue to give Lupin the opportunity to prove its case, and Lupin did not.

The evidence adduced at trial shows that Tibotec selected darunavir as a compound of interest based on its own proprietary data and research, not from any publicly available literature. The Court finds that Lupin has not provided clear and convincing evidence of anything in the prior art which would have made darunavir a compound of interest. With that failure, Lupin's obviousness case fails.

By all rights, the analysis of Lupin's five-step "Obviousness Pathway" ends. Lupin failed to establish the prerequisite— the sine qua non—that darunavir was a compound of interest in 2002. This path has no beginning. To repeat, passage is ended before begun. Lupin is attempting to leap over its shadow, a futile effort. Even briefly surveying the remainder of the proffered pathway leads the Court to conclude that steps 2–5 are meritless. Particularly so since Lupin's main pathway guide, its expert Dr. Zaworotko, at trial expressed a material opinion diametrically opposed to one he had earlier published. Such contradiction was not inadvertent. Specifically, on cross-examination, Dr. Zaworotko stated: "the only data I recall was that darunavir crystallizes from isopropanol very readily.... I've seen no evidence that suggests that darunavir is hard to crystallize. Seen no data, experimental data." Zaworotko Tr. 1715:25–1716:11. Dr. Zaworotko was then shown a chapter in a book he had edited that specifically describes darunavir ethanolate as "unusual." See PTX 842A (Organic Crystal Engineering) at 75. Upon questioning from this Court, Dr. Zaworotko conceded that he agreed with the paragraph containing that assertion "in its entirety." Zaworotko Tr. 1717:18–1719:5. In so doing, Dr. Zaworotko agreed that darunavir ethanolate is "an unusual crystal form," one of "very few" ethanolates on the market and that a skilled artisan would have had an "aversion" to attempting to develop an ethanolate of darunavir because of the many difficulties and drawbacks to be expected. Specifically, one of skill would have no reasonable expectation of success in developing an ethanolate form suitable for pharmaceutical composition, as required by claim 4, because solvates are "prone to solvent loss" on standing and with it, a "loss of chemical stability." PTX 842A (Organic Crystal Engineering) at 75. Darunavir ethanolate is one of "very few" ethanolates, and its discovery could not have been expected. Zaworotko Tr. 1717:18–1719:5 (agreeing with the contents of the paragraph in PTX 842A at 75); see also Myerson Tr. 1825:17–25. The advice of Defendant's expert as a pathway guide is not credible.

4. Secondary Considerations

a. Commercial success

Janssen again supports this secondary consideration by asserting that Prezista is "the number-one selling protease inhibitor," Leffler Tr. 2311:22–2312:3, with worldwide sales in 2013 of more than $1.6 billion and U.S. sales of more than $800 million. Id. at 2288:21–24, 2311:9–15; Stoffels Tr. 112:12–14.

And again, Lupin argues that there is no nexus between the alleged evidence of secondary considerations and the '645 Patent. Lupin puts forth the same arguments as to the '645 Patent with regard to nexus as it did for the '015 Patent (i.e., physicians do not prescribe Prezista based on the fact that the darunavir active pharmaceutical ingredient is in the ethanolate form; there is no clinical evidence that attributes the benefits of Prezista to the ethanol solvate of darunavir as opposed to the darunavir molecule itself; Prezista is always used in combination therapy, etc.). Again, this Court is unpersuaded by Lupin's arguments.

Prezista is a pharmaceutical composition of a 1:1 ethanolate form of darunavir, exactly as claimed in the '645 Patent. See DTX 623 (Highlights of Prescribing Information for Prezista ) at PREZ05142289 (describing Prezista as "[darunavir ] monoethanolate"). That Prezista is exactly what is claimed "trigger[s] a presumption of nexus between the drug's success and the claimed invention ..." Teva Pharms. USA, Inc. v. Sandoz, Inc., 723 F.3d 1363, 1372 (Fed.Cir.2013). Lupin has not overcome that presumption. It is undisputed that the formulation of a molecule "can have a significant impact on ... the [bio]availability of [a] drug compound....," Stoffels Tr. 143:7–10, and that the formulation of Prezista"[a]bsolutely" matters for its success, id. at 102:16–17, and has been "important" to that success, Myerson Tr. 1832:5–1833:1. It is quite logical that a drug's formulation can be critical to its commercial success: "You have to deliver [a drug] molecule to people, and you have to deliver it in a way that they comply, that they take it every day, and you have to deliver it in a way that you can manufacture." Id. at 1832:13–1833:1. "[Y]ou can have a good molecule, and if you can't formulate it, it's not going to be very useful, or if you formulate it in a way that people don't want to take, it will be less useful." Id.

The record provides examples of AIDS drugs that did not succeed because of formulation problems. See Zingman Tr. 850:7–25 (discussing amprenavir's poor bioavailability, poor solubility and high pill burden); Myerson Tr. 1833:6–11 (same); PTX 282 ('989 Patent) at col. 1:32–54 (same); Stoffels Tr. 142:7–143:25, 102:9–15, 81:23–83:5 (discussing Intelence's "high pill burden"); Zingman Tr. 847:11–18 (discussing ritonavir's instability); Myerson Tr. 1833:12–21 (same); Zaworotko Tr. 1601:14–1602:1. In contrast to these examples, Prezista is available in a stable, bioavailable formulation that allows for once-daily dosing for some categories of patients and twice daily dosing for others. DTX 623 (Highlights of Prescribing Information for Prezista ) at PREZ05142265. This favorable dosage regimen has contributed to Prezista's success. Falcon Tr. 2487:10–12, 2485:8–10.

Prezista has only been sold in the ethanolate formulation that is claimed in the '645 Patent. Wigerinck Tr. 335:19–336:5. The '645 invention is "an inextricable and essential part of what doctors are prescribing" when they prescribe Prezista. Teva Pharms. USA, Inc. v. Sandoz, Inc., 876 F.Supp.2d 295, 416–17 (S.D.N.Y.2012), aff'd, 723 F.3d 1363 (Fed.Cir.2013), cert. granted on other issues, ––– U.S. ––––, 134 S.Ct. 1761, 188 L.Ed.2d 592 (2014). Prezista's success "is attributable to the medical benefits of its active ingredient, including its method of administration, [and] bioavailability...." Roche Palo Alto LLC v. Ranbaxy Labs., Ltd., No. 06–2003, 2009 WL 3261252, at *27 (D.N.J. Sept. 30, 2009).

The Court finds that the secondary consideration of commercial success weighs against a finding of obviousness.

b. Unexpected Results

Janssen asserts that before the '645 invention, persons of ordinary skill in the art would not have expected a solvate of darunavir to be suitable for use as a pharmaceutical composition. "[G]enerally the pharmaceutical industry always prefers to use non-solvated forms where possible." Myerson Tr. 584:16–19. "[S]olvates can be prone to solvent loss," leading to "undesirable changes" including "loss of chemical stability." Zaworotko Tr. 1717:18–1719:5 (agreeing with PTX 842A (Organic Crystal Engineering) at 75). This helps explain the "aversion" to using solvates in pharmaceutical products and is one reason why there are "very few examples ... [of] actual solvates in approved products." Id. (agreeing with PTX 842A (Organic Crystal Engineering) at 75).

Janssen maintains that darunavir ethanolate has unexpected properties that enable its formulation as a drug product, in particular its "stability, particularly under heat" and "a very high desolvation temperature," that are "really unexpected" for solvates in general and for channel solvates in particular. Myerson Tr. 1836:24–1837:7. Janssen repeats that darunavir ethanolate has unexpectedly good bioavailability. In seeking approval for its ANDA products, Mylan represented to the FDA that its amorphous darunavir is bioequivalent to darunavir ethanolate. Myerson Tr. 1847:8–11; DTX 1127 (Mylan's Request for Waiver of In-vivo BA/BE Studies). This is unexpected because "generally, amorphous forms have higher solubility and dissolution rate than crystalline forms," which means they generally demonstrate better bioavailability. Myerson Tr. 1847:12–14.

Lupin argues that as of May 2002, it would have been reasonable for a person of ordinary skill in the art looking to crystallize darunavir from ethanol to expect that a darunavir ethanolate could form. Zaworotko Tr. 1633:4–15, 1644:16–22. Lupin also contends that as of 2002, a person of ordinary skill in the art would have arrived at the reasonable expectation by reviewing the published literature and observing that most ethanolates occur in a 1:1 ratio. Zaworotko Tr. 1634:15–1635:16. Lupin also claims that it would not have been surprising that different darunavir drug materials can produce medicines that are bioequivalent. Zaworotko Tr. 1640:21–23. As support, Lupin argues that the differences in properties among different solvates and polymorphs of a molecule—including solubility and bioavailability—are minor and would not impact the performance of the medicine. Id. at 1640:21–1641:1, 1641:12–21, 1728:23–1729:21.

This Court finds that Janssen's arguments regarding unexpected results are persuasive and adopts its reasoning to conclude that the secondary consideration of unexpected results weighs against a finding of obviousness. Lupin's argument, based on Dr. Zaworotko's testimony that it would not have been surprising for a channel solvate to perform well in a pharmaceutical formulation, is unconvincing. This Court finds Dr. Zaworotko's testimony has no credibility as to this issue given his general agreement with statements to the contrary in Organic Crystal Engineering. See Zaworotko Tr. 1717:18–1719:5 (agreeing with PTX 842A (Organic Crystal Engineering) at 75).

c. Praise and industry acceptance

Again, Janssen asserts that Prezista has received widespread praise, including the highest possible recommendation in the guidelines promulgated by the U.S. Department of Health and Human Services, and its industry acceptance is reflected by the fact that it is the leading protease inhibitor.

Lupin puts forth the same lack of nexus arguments as it did with regard to commercial success. For the same reasons discussed earlier, the Court finds a causal nexus between claim 4 of the '645 Patent and the praise and industry acceptance of Prezista, and that these secondary factors also weigh against a finding of obviousness.

d. Copying

Janssen asserts that Lupin copied the '645 invention. Myerson Tr. 1847:15–24.

Again Lupin responds that in the context of Hatch–Waxman cases, the Federal Circuit has deemed evidence of copying by generic manufacturers to be "not probative of nonobviousness because a showing of bioequivalence is required for FDA approval." Bayer, 713 F.3d at 1377. Again Lupin is correct, and as such the Court does not consider this secondary consideration of nonobviousness.

5. Conclusion: The '645 Patent was not obvious

Lupin has not met its burden to prove by clear and convincing evidence that the invention of claim 4 of the '645 Patent would have been obvious to a person of ordinary skill in the art. Lupin has failed to demonstrate by clear and convincing evidence that a person of ordinary skill in the art would have selected darunavir as a lead compound or compound of interest for formulation. It follows then that the remaining steps of Lupin's "Obviousness Pathway" are substantively irrelevant.

The relevant secondary considerations, commercial success, unexpected results, praise and industry acceptance, all weigh against a finding of obviousness.

Considering the Graham factors, this Court finds that Lupin has not met its burden to demonstrate by clear and convincing evidence that the invention of claim 4 of the '645 Patent would have been obvious to a person of ordinary skill in the art.

ii. Enablement and Written Description under 35 U.S.C. § 112

Lupin's expert Dr. Philip Gould raised enablement and written description defenses for the '645 Patent. These defenses for the '645 Patent are directed at the "carrier" limitations in claim 4, which recite that the "inert carrier is a pharmaceutically acceptable carrier." PTX 1 ('645 Patent) at col. 30:36–42.

On direct examination, in testimony presumably aimed at Lupin's obviousness defense, Dr. Gould testified that the '645 Patent's specification provides examples of "suitable inert carriers." Gould Tr. 1276:5–9; see PTX 1 ('645 Patent) at col. 14:55–58. He also agreed that inert carriers for pharmaceutical compositions were known in the prior art, including in the ' 775 Patent, which discloses and claims the darunavir compound, and identifies acceptable inert carriers for the compounds it discloses. Gould Tr. 1271:20–1272:18; see DTX 5 ('775 Patent) at col. 217:36–44. Dr. Gould stated that nothing more than "[r]outine experimentation" would be needed for a person of ordinary skill in the art to "prepare the claimed composition that contains an ethanol solvate of the drug and carrier." Gould Tr. 1271:10–15. He also said that the disclosures in the patent and the prior art would have given a person of ordinary skill "a reasonable likelihood of successfully using an inert carrier with an ethanol solvate form of [darunavir ] as disclosed in claim 4 of the '645 patent...." Id. at 1272:19–25. Dr. Gould asserted that these opinions "wouldn't change" when claim 4's additional requirement that "the inert carrier is a pharmaceutically acceptable ... carrier" is taken into account. Id. at 1273:1–6.

After eliciting that testimony from Dr. Gould, Lupin's counsel then asked him about Lupin's invalidity defense under section 112. Gould Tr. 1273:7–8. At that point, Dr. Gould offered opinions that are contradicted by the opinions discussed above. He opined that the patent and prior art do not provide sufficient information to allow a person of ordinary skill to select and make an inert carrier that would provide "special formulation properties" and "positive influences" such as good bioavailability—which, as he previously testified, are not requirements of the claim, id. at 1271:3–5–and that claim 4 accordingly is invalid for failure to meet the enablement and written description requirements. Id. at 1273:23–1278:2.

This Court finds that the internal contradictions in Dr. Gould's testimony raise serious concerns about his credibility as a witness. But even apart from those concerns, his opinions on enablement and written description cannot withstand scrutiny. Dr. Gould was correct when he testified that claim 4 does not require any "formulation properties" or "desired outcomes" for the "inert carrier" or the claimed "composition." Id. at 1271:3–5. The carrier only needs to be "inert" and "pharmaceutically acceptable." PTX 1 ('645 Patent) at col. 30:36–42. Nothing more is required. In the absence of such requirements in the claim, the specification's disclosure of "[e]xamples of suitable inert carriers," PTX 1 at col. 14:55–58, together with the knowledge of persons skilled in the art, is more than sufficient to satisfy the enablement requirement. The patent adequately describes and enables these claim limitations. As Dr. Myerson credibly testified, "the patent discloses the ethanolate solvate in a ratio of one to one. It discloses pharmaceutically acceptable inert carriers. If one of ordinary skill wanted to, they could just take the inert carrier and the active ingredient and fill them into a capsule and they'd have a pharmaceutical formulation to make the claim." Myerson Tr. 1829:4–24. In the initial part of his testimony, Dr. Gould agreed. Gould Tr. 1258:7–10, 1276:5–9, 1271:10–1273:6.

Dr. Gould's testimony on obviousness is sufficient to defeat his testimony on enablement and written description. Dr. Gould explained that a person of ordinary skill in the art would know how to use the "suitable carriers" identified in the '645 Patent specification in a pharmaceutical formulation, Gould Tr. 1294:25–1295:5; that "suitable pharmaceutical carriers or excipients as disclosed in the art are very common and frequently used by formulators for direct compression products," id. at 1295:6–10; and that "the preparation of tablets using these excipients are well described in standard texts which teach their suitability for tablet formulation and give standard ranges of these materials that are frequently employed," id. at 1295:13–19. He also agreed that "well used references," such as the Remington and Martindale references, teach the use of inert carriers. Id. at 1296:14–21. Consistent with Dr. Gould's testimony, the specification provides all the detail one of skill in the art would need to practice the invention and to demonstrate that the inventors were in possession of the claimed invention. As such, on the evidence presented, Lupin has not met its burden of proving by clear and convincing evidence that claim 4 of the '645 Patent is invalid for failure to satisfy the enablement or written description requirements.

III. Remedies

Janssen seeks three types of relief in this case. With regard to all patents-in-suit, Janssen seeks a declaration that Defendants' ANDA products infringe (or will infringe) valid claims of the asserted patents. Pre–Trial Order at 15–17. As to the '015 and '411 Patents, the infringement has not yet occurred and Janssen seeks a declaratory judgment under the Declaratory Judgments Act, 28 U.S.C. 2201(a). As to the '645 Patent, Lupin's filing of its ANDA constitutes an act of infringement under the Hatch–Waxman Act and Janssen seeks both a judgment of infringement and declaratory relief.

Second, with respect to all patents-in-suit, Janssen seeks a permanent injunction barring Defendants from importing into the United States their ANDA products and selling them. Pre–Trial Order at 15–17.

Third, as to the '645 Patent, Janssen seeks an order under 35 U.S.C. § 271(e)(4)(A) that the effective date of any FDA approval of Lupin's ANDA be no earlier than the day after the expiration of the '645 Patent and any associated exclusivity. Id. at 16.

At the close of Janssen's case-in-chief, Defendants moved for judgment under Federal Rule of Civil Procedure 52(c), arguing that Janssen had failed to show it was entitled to any permanent injunction. Tr. 2269:1–2270:21. The Court reserved judgment on the motion, Tr. 2273:12, and directed the parties to submit briefing on the issue, Tr. 2358:5–9, which the parties did, see ECF Nos. 840, 841. Defendants' Rule 52(c) motion overlaps with the issues the Court must decide regarding Janssen's entitlement to the permanent injunction remedies it seeks. As such, the Court's rulings regarding permanent injunctions also resolve the 52(c) motion. As will be seen, Defendants' Rule 52(c) motion is denied.

A. The'015 Patent

i. Declaratory Judgment that the '015 Patent is Not Invalid and is Infringed by Lupin

This Court previously granted summary judgment to Plaintiffs that the importation, offer to sell, sale, or use of Lupin's ANDA Products would infringe claim 1 of the '015 Patent under 35 U.S.C. § 271(g). Summ. J. Op. at 49–59. At trial, Lupin failed to prove by clear and convincing evidence that claim 1 of the '015 Patent is invalid as obvious under 35 U.S.C. § 103, see supra Section II(B)(i)(5) or invalid as non-enabled under 35 U.S.C. § 112, ¶ 1, see supra Section II(B)(ii).

As a consequence, Janssen is entitled to a declaratory judgment that claim 1 of the '015 Patent is not invalid and that importing, selling, offering to sell, or using the generic darunavir tablets described in Lupin's ANDA would constitute infringement of the '015 Patent under 35 U.S.C. § 271(g). See Pre–Trial Order at 16.

ii. Permanent Injunction Barring Lupin from Importing or Selling its ANDA Products Before the '015 Patent Expires

Janssen also seeks a permanent injunction enjoining Lupin from commercially manufacturing, selling or offering for sale, using, or importing the generic darunavir tablets described in Lupin's ANDA in the United States, or any darunavir product that includes a bis-THF component made by any colorable variation of the process used to make Lupin's ANDA products, until after the expiration date of the '015 Patent. Pre–Trial Order at 17.

In eBay Inc. v. MercExchange, LLC the Supreme Court confirmed that the traditional four factor test for obtaining a permanent injunction applies in this context. 547 U.S. 388, 391, 126 S.Ct. 1837, 164 L.Ed.2d 641 (2006). In doing so, the Court rejected "categorical rule[s]" concerning the granting or denying of injunctive relief, and held that the "decision whether to grant or deny injunctive relief rests within the equitable discretion of the district courts." Id. at 393, 394, 126 S.Ct. 1837. Here Janssen must demonstrate

(1) that it has suffered an irreparable injury; (2) that remedies available at law, such as monetary damages, are inadequate to compensate for that injury; (3) that, considering the balance of hardships between the plaintiff and defendant, a remedy in equity is warranted; and (4) that the public interest would not be disserved by a permanent injunction.

Id. at 391, 126 S.Ct. 1837. Lupin makes a series of arguments that Janssen has failed to establish that it is entitled to a permanent injunction based on Lupin's infringement of the '015 Patent.⁵ The Court concludes, however, that Janssen has satisfied its burden and that an injunction is warranted.

5 Mylan's and Lupin's arguments regarding Janssen's entitlement to permanent injunctive relief on the three patents at issue substantially overlap, so the Court sets out its analysis in greatest detail here, and will refer to this section in its analysis of the '411 and '645 Patents. In addition, Mylan's and Lupin's arguments in the parties' joint proposed findings are the same as those in their Rule 52(c) motion. The Court refers to those made in the joint proposed findings for ease of reference.

1. Irreparable Injury

After eBay, "there is no presumption of irreparable harm upon a finding of patent infringement." Apple Inc. v. Samsung Elecs. Co. (Apple III ), 735 F.3d 1352, 1359 (Fed.Cir.2013). "[T]o satisfy the irreparable harm factor in a patent infringement suit, a patentee must establish both of the following requirements: 1) that absent an injunction, it will suffer irreparable harm, and 2) that a sufficiently strong causal nexus relates the alleged harm to the alleged infringement." Apple Inc. v. Samsung Elecs. Co. (Apple II ), 695 F.3d 1370, 1374 (Fed.Cir.2012). The "irreparable harm and causal nexus inquiries" are "inextricably related concepts," but they "may be separated for the ease of analysis." Id. The Federal Circuit has now made clear that "the causal nexus requirement is part of the irreparable harm factor. Without a showing of causal nexus, there is no relevant irreparable harm. In other words, there cannot be one without the other." Apple III, 735 F.3d at 1363. As suggested by the Circuit, the Court addresses each aspect of the irreparable injury inquiry separately.

a. Irreparable Harm

"Irreparable injury encompasses different types of losses that are often difficult to quantify...." Douglas Dynamics, LLC v. Buyers Products Co., 717 F.3d 1336, 1344 (Fed.Cir.2013). "Price erosion, loss of goodwill, damage to reputation, and loss of business opportunities are all valid grounds for finding irreparable harm." Celsis In Vitro, Inc. v. CellzDirect, Inc., 664 F.3d 922, 930–31 (Fed.Cir.2012). A showing of lost market share and sales can support a finding of irreparable harm. See Robert Bosch LLC v. Pylon Mfg. Corp., 659 F.3d 1142, 1154 (Fed.Cir.2011) ; i4i Ltd. P'ship v. Microsoft Corp., 598 F.3d 831, 861–62 (Fed.Cir.2010), aff'd, 564 U.S. 91, 131 S.Ct. 2238, 180 L.Ed.2d 131 (2011). A patentee can also demonstrate irreparable harm from a diminished ability to invest in research and development. See Bio–Tech. Gen. Corp. v. Genentech, Inc., 80 F.3d 1553, 1566 (Fed.Cir.1996) ; AstraZeneca LP v. Apotex, Inc., 623 F.Supp.2d 579, 612–13 (D.N.J.2009), supplemented, 623 F.Supp.2d 615 (D.N.J.2009), aff'd, 633 F.3d 1042 (Fed.Cir.2010). In addition, "[w]here two companies are in competition against one another, the patentee suffers the harm—often irreparable—of being forced to compete against products that incorporate and infringe its own patented inventions." Douglas Dynamics, 717 F.3d at 1345 ; see also Presidio Components, Inc. v. Am. Technical Ceramics Corp., 702 F.3d 1351, 1363 (Fed.Cir.2012) ("Direct competition in the same market is certainly one factor suggesting strongly the potential for irreparable harm without enforcement of the right to exclude." (citation omitted)).

The Court finds that Janssen has demonstrated that it will suffer irreparable harm from Lupin's infringement in the form of lost sales, price erosion, research and development ability, and having to compete directly with an infringing competitor.

With respect to lost sales, Defendants' economics expert, Dr. Leffler, stated that Prezista prescriptions in the United States generated over $800 million last year. Leffler Tr. 2310:6–15. Janssen presented the testimony of Jessica Martori, Lupin's senior manager of portfolio development, regarding "the projected market share or market potential of Lupin's ANDA products." Martori Tr. 2256:7–22. Ms. Martori testified that in the first year of the launch of its generic product, 98 percent of the market would be generic, followed by 100 percent every year thereafter. Id. at 2259:21–25. Janssen has clearly demonstrated that it would lose virtually all, if not all, of its sales to Lupin. Accord Automated Merch. Sys., Inc. v. Crane Co., 357 Fed.Appx. 297, 301 (Fed.Cir.2009) ("To the extent that failing to grant a preliminary injunction would permit Crane to drop its prices in order to drive AMS out of the market entirely, this might support a finding of irreparable harm sufficient to warrant a preliminary injunction.").

6 ⁶ Any objection by Lupin that the [redacted] testimony show that Janssen will lose this amount of sales to a generic, but not necessarily to Lupin, is misplaced. The testimony shows that sales will be lost upon generic entry, which supports the conclusion that the same loss would occur if Lupin were the only generic to enter. The happenstance that other generics might also enter does not change this conclusion. See Robert Bosch LLC, 659 F.3d at 1154 ("While the party seeking an injunction bears the burden of showing lost market share, this showing need not be made with direct evidence") [redacted]

There is also ample evidence that entry by Lupin would lead to price erosion [redacted] See also Martori Tr. 2261:6–7 ("Because once the generics launch, they take a majority of the brand business away, and they lower price.").

In addition, Janssen presented the testimony of Dr. Stoffels, the chief scientific officer of Johnson & Johnson, who is responsible for all product development. Stoffels Tr. 63:3–65:25. Dr. Stoffels testified that out of every hundred dollars of Prezista sales, twenty-two are reinvested into new research and development. Id. at 115:22–116:1. Based on his estimate that Janssen would lose 70 to 80 percent of its sales upon a generic launch, Dr. Stoffels stated there would be a "few hundred million dollars less to be invested in R & D for new drugs and new products." Id. at 116:12–14. This testimony is unrebutted, and the Court finds that the loss represented by 22 percent of Prezista sales is a significant harm to Janssen's research and development efforts.

Finally, there is no question that Lupin would be Janssen's direct competitor upon launching its generic products, which the Court has already found would be made by a process that would infringe Janssen's valid patent. This also supports a finding of irreparable harm. Douglas Dynamics, 717 F.3d at 1345 ; Presidio Components, 702 F.3d at 1363 ; see also Trebro Mfg., Inc. v. Firefly Equip., LLC, 748 F.3d 1159, 1171 (Fed.Cir.2014) ("Trebro and FireFly are direct competitors selling competing products in this market. Thus, the record strongly shows a probability for irreparable harm.").

The Court finds that these four circumstances—lost sales, price erosion, diminished research and development assets, and having to compete directly with an entity selling a product made by an infringing process—together demonstrate that Janssen will be irreparably harmed absent an injunction. As next explained, there is a causal nexus between the irreparable harm and Lupin's infringement.

b. Causal Nexus

The Federal Circuit has recently emphasized the causal nexus aspect of the irreparable injury inquiry in a series of appeals between Apple, Inc. and Samsung Electronics Co., Ltd. See Apple Inc. v. Samsung Elecs. Co. (Apple I ), 678 F.3d 1314 (Fed.Cir.2012) ; Apple II ; Apple III . "[T]he causal nexus requirement is simply a way of distinguishing between irreparable harm caused by patent infringement and irreparable harm caused by otherwise lawful competition.... The former type of harm may weigh in favor of an injunction, whereas the latter does not." Apple III, 735 F.3d at 1361. "[T]he purpose of the causal nexus requirement is to show that the patentee is irreparably harmed by the infringement. " Id. at 1363. This requirement "reflects general tort principles of causation and applies equally to the preliminary and permanent injunction contexts." Id. at 1361.

1) The Apple Cases

The Apple cases involved whether Apple was entitled to an injunction based on Samsung's infringement of a select few patents incorporated into "complex, multi-featured smartphones and tablets." Id. at 1363. In this context, the court explained:

[I]t may very well be that the accused product would sell almost as well without incorporating the patented feature. And in that case, even if the competitive injury that results from selling the accused device is substantial, the harm that flows from the alleged infringement (the only harm that should count) is not.... [T]he causal nexus inquiry ... informs whether the patentee's allegations of irreparable harm are pertinent to the injunctive relief analysis, or whether the patentee seeks to leverage its patent for competitive gain beyond that which the inventive contribution and value of the patent warrant.... The relevant question is not whether there is some causal relationship between the asserted injury and the infringing conduct, but to what extent the harm resulting from selling the accused product can be ascribed to the infringement.... The patentee must rather show that the infringing feature drives consumer demand for the accused product.

Apple II, 695 F.3d at 1374–75. The Circuit also discussed other recent decisions in which it had not addressed the causal nexus requirement in detail. Those cases "involved relatively simple products—at least in the sense that they had a small number of features when compared to the complex, multi-featured smartphones and tablets at issue." Apple III, 735 F.3d at 1362. The court made clear that "the causal nexus requirement applies regardless of the complexity of the products. It just may be more easily satisfied (indeed, perhaps even conceded) for relatively ‘simple’ products." Id.

The Circuit made two additional points. The causal nexus inquiry did not mean "Apple must show that a patented feature is the one and only reason for consumer demand.... [R]ather than show that a patented feature is the exclusive reason for consumer demand, Apple must show some connection between the patented feature and demand for Samsung's products." Id. at 1364. The court stressed that there "might be a variety of ways to make this required showing," such as "with evidence that a patented feature is one of several features that cause consumers to make their purchasing decisions," or that the inclusion or exclusion of a patented feature makes a product significantly more or less desirable. Id. As example:

[A] battery does not necessarily drive demand for a laptop computer simply because its removal would render the laptop ineffective as a portable computer. That is because consumers often do not choose a laptop based on its battery, and presumably at this point, no inventor has a patent covering all laptop batteries. Nevertheless, it is indisputable that the ability to carry around a computer without having to plug it in is one of the reasons people buy laptops. Thus, if the first person to invent a laptop battery had obtained a patent covering all laptop batteries, then it would be reasonable to say that the patented invention was a driver of demand for laptops.

Id. (citation omitted).

Second, the court acknowledged "there may be circumstances where it is logical and equitable to view patents in the aggregate," such as "where they all relate to the same technology or where they combine to make a product significantly more valuable." Id. at 1365.

2) Discussion

Janssen and Lupin argue at length over whether Janssen has demonstrated a causal nexus between the irreparable harm and Lupin's infringement. Lupin stresses that Janssen has failed to meet the standard articulated in the Apple cases. The Court disagrees, and finds that there is a causal nexus between the irreparable harm and infringement.

At the outset, the Court rejects an attempt by Janssen to cut the inquiry short. Janssen argues that "Congress anticipated and addressed the nexus issue" in 35 U.S.C. § 271(g). Janssen's Br. in Opp'n to Defs.' Mot. for J. on Partial Findings ("Janssen's 52(c) Br.") at 16 (ECF No. 840). That section provides, "Whoever without authority imports into the United States or offers to sell, sells, or uses within the United States a product which is made by a process patented in the United States shall be liable as an infringer." 35 U.S.C. § 271(g). There are two exceptions: a product that is "materially changed by subsequent processes" and a product that "becomes a trivial and nonessential component of another product." § 271(g)(1), (2). Janssen argues that the section's legislative history shows that the exceptions "restrict[ ] the scope of the bill to exclude ultimate products that, because of intervening manufacturing steps, cease to have a reasonable nexus with the patented process." Janssen's 52(c) Br. at 16 (quoting S.Rep. No. 100–83, at 36 (1987) (Janssen's emphasis) (quotation marks omitted)). Janssen's argument is that unless one of the exceptions applies, a causal nexus exists. And because this Court found in its summary judgment opinion that neither exception applied, see Summ. J. Op. at 54–59, there is a causal nexus in this case.

The Court disagrees. Even giving the one sentence of legislative history the weight Janssen does, the Court fails to see how this resolves the causal nexus inquiry. That section and sentence focus on whether the final imported product is related closely enough to the patented process to justify a finding of infringement. It has nothing to do with the question of remedy if infringement is found. The Supreme Court emphasized in eBay that "the creation of a right is distinct from the provision of remedies for violations of that right." 547 U.S. at 392, 126 S.Ct. 1837. The causal nexus inquiry focuses on the relationship between irreparable harm and infringement, a separate issue. If the Court adopted Janssen's reasoning, then in every infringement case under section 271(g) the causal nexus requirement would be satisfied, moving trial courts back toward the presumption of injunctive relief that the Supreme Court rejected in eBay.

Returning to our inquiry, the Court finds Janssen has demonstrated a causal nexus between Lupin's infringement and the irreparable harm. Lupin's process for making its generic darunavir products infringes the '015 Patent. Summ. J. Op. at 59. The sale of those products would lead directly to the irreparable harms shown—lost sales (see Martori Tr. 2259:21–25), price erosion [redacted] adverse impact on research and development (see Stoffels Tr. 115:22–116:1), and the need to compete directly with an infringing competitor. Without infringing the '015 Patent, Lupin would not be able to make the products proposed in its ANDA. And without an injunction, Lupin would use Janssen's patented process to produce a product that would substantially destroy Janssen's Prezista business.

Lupin attempts to draw parallels between this case and the Apple cases to claim that Janssen has not shown a causal nexus. Lupin argues that Janssen "will lose sales if Lupin comes on the market irrespective of whether Lupin's process for making bis-THF ... did or did not use the '015 patent's process steps." Joint Proposed Findings of Fact ¶ D1118. Lupin points to testimony that doctors and patients care about getting the darunavir molecule into the body, about the number of pills patients have to take, and about the method of using Prezista. Zingman Tr. 861:9–l 5, 861:25–862:5, 889:25–890:12; Leffler Tr. 2282:10–17, 2284:9–14; Falcon Tr. 2485:8–13. There is also testimony that "there's no evidence that the specific processes in [the '015 and '411] patents is key to the effectiveness of a darunavir —containing combination regimen." Zingman Tr. 862:3–5. Lupin says that Janssen "failed to show (e.g., through a customer survey or comparable market research) that anyone would actually buy Lupin Ltd.'s ANDA products" due to Lupin's use of the '015 Patent's process. Joint Proposed Findings of Fact ¶ D1124. Lupin concludes that Janssen has "put forth no evidence that the claimed, non-prior art elements of the '015 patent are what drive sales or consumer demand." Id. ¶ 1126 (emphasis omitted).

First, this case is unlike the Apple cases because it does not involve "complex, multi-featured" products. Rather, the Court finds the darunavir products at issue more akin to the "relatively simple products" the Federal Circuit contrasted with the smartphones and tablets in the Apple cases. ⁷See Apple III, 735 F.3d at 1362. Indeed, Lupin states that "Prezista contains many desirable features," Joint Proposed Findings of Fact ¶ D1120, but lists only three. By contrast, the several patented features at issue in the Apple cases were a "small minority" of the large number of features in those products. Apple II, 695 F.3d at 1374. Here this means that the causal nexus inquiry "may be more easily satisfied." Apple III, 735 F.3d at 1362.

7 "Simple" and "complex" refer to the number of features the product has.

Lupin's comparison to the Apple cases is also misleading in its use of the term "features." In the Apple cases, the "features" were attributes of the finished consumer good; they did not concern the process the companies used to make that good. Calling the '015 process a "feature" of the finished darunavir products is inaccurate. This leads to the main problem with Lupin's argument: unlike the case of various features of a finished consumer good, it is not possible to separate the process for making the finished darunavir product from the product itself in evaluating consumer demand and nexus. The Apple cases did not deal with a claim of patent infringement of a process to make tablets and smartphones. Were it necessary to show that consumers bought a product because of the process by which it is made, there would be a de facto result of no irreparable harm and no permanent injunction in patent infringement cases concerning processes. But such a result would run counter to eBay, 547 U.S. at 393, 394, 126 S.Ct. 1837. And that has not been so in past process patent cases. See, e.g., Amgen Inc. v. F. Hoffman–La Roche Ltd., 580 F.3d 1340, 1386 (Fed.Cir.2009) (affirming injunction in part based on finding of infringement of process patents under 35 U.S.C. § 271(g) ); Kemin Foods, L.C. v. Pigmentos Vegetales Del Centro S.A. de C.V., 464 F.3d 1339, 1356 (Fed.Cir.2006) (affirming district court's award of injunction for infringement of patented process); accord eBay, 547 U.S. at 395, 126 S.Ct. 1837 (Roberts, C.J., concurring) (noting the "long tradition of equity practice" of courts granting "injunctive relief upon a finding of infringement in the vast majority of patent cases" because of "the difficulty of protecting a right to exclude through monetary remedies that allow an infringer to use an invention against the patentee's wishes," and explaining that "like cases should be decided alike" and that in exercising its equitable discretion, a district court should remember that "a page of history is worth a volume of logic." (citations and quotation marks omitted)); id. at 396, 126 S.Ct. 1837 (Kennedy, J., concurring) ("To the extent earlier cases establish a pattern of granting an injunction against patent infringers almost as a matter of course, this pattern simply illustrates the result of the four-factor test in the contexts then prevalent. The lesson of the historical practice, therefore, is most helpful and instructive when the circumstances of a case bear substantial parallels to litigation the courts have confronted before."). The Court finds this distinction crucial, and does not accept Lupin's comparison of the '015 process with the features at issue in Apple.

8 ⁸ Lupin makes two additional arguments. Because Janssen "will purportedly lose sales to Mylan as readily as to Lupin, ... it is plainly not any process patented in the '015 Patent that drives consumer demand." Joint Proposed Findings of Fact ¶ D1135. Second, that the ' 015 Patent allows for commercial scale production of bis-THF "does not provide a nexus hook" for the irreparable harm. Id. ¶ D1138. Both of these arguments are based on the premise the Court rejects—that it is reasonable to equate the process at issue here with the features at issue in the Apple cases.

In addition, the Federal Circuit has made clear that the potential for the complete destruction of a patentee's business with respect to the patented product can make a difference in the causal nexus inquiry. See Apple I, 678 F.3d at 1324 ("The narrow injunction upheld by this court [in a prior case] served only to protect the patented product from obsolescence.... Here, in contrast, the district court found that the alleged acts of infringement do not threaten to have any such dramatic effects on the market generally or on Apple's share of that market."). Here, as the Court has already found, Lupin's infringement would lead to the complete, or nearly complete, destruction of Janssen's Prezista business. See Martori Tr. 2259:21–25; [redacted]

Lupin makes an additional argument that Janssen has "failed to show which entity is actually selling in the United States, which corporate entity will lose sales in the United States; or that any person or entity who is a Plaintiff here can be deemed synonymous with the patent holder." Joint Proposed Findings of Fact ¶ D1160. Lupin argues that, as a result, Janssen has not demonstrated that any entity has standing or will suffer any irreparable harm. Id. ¶ D1161. But Lupin and Janssen stipulated this issue out of the case in the Pre–Trial Order, which provides that "Janssen," defined as both Janssen R & D Ireland and Janssen Products, L.P., "sells" Prezista. Pre–Trial Order at 2 § 1; 9 ¶ 14. This valid stipulation moots such argument. See Waldorf v. Shuta, 142 F.3d 601, 616 (3d Cir.1998) ("[I]t is a well-recognized rule of law that valid stipulations entered into freely and fairly, and approved by the court, should not be lightly set aside." (citation and quotation marks omitted)).

The Court finds that Janssen has shown irreparable harm that is causally related to Lupin's infringement of the '015 Patent, which weighs in favor of a permanent injunction.

2. Inadequacy of Legal Remedies

Janssen must demonstrate that " ‘remedies available at law, such as monetary damages, are inadequate to compensate’ the patentee for the irreparable harm it has suffered." Apple III, 735 F.3d at 1368 (quoting eBay, 547 U.S. at 391, 126 S.Ct. 1837 ). The Court finds that monetary damages would be inadequate to remedy the harms Janssen will suffer.

The Court notes that a finding of irreparable injury will, in most cases, go hand in hand with a finding that legal remedies are inadequate. See Nicholas P. Chan, Balancing Judicial Misvaluation and Patent Hold–Up: Some Principles for Considering Injunctive Relief After eBay, 59 UCLA L.Rev. 746, 783 (2012) ("Most courts consider the first two factors interchangeable, and they are usually analyzed together: An injury is irreparable when monetary damages are inadequate to compensate for that injury."). This is such a case.

The record shows that Janssen will not only lose sales and be forced to compete with a direct competitor, but that the competition with its direct competitor will lead directly to the destruction of its Prezista business. Monetary compensation can alleviate some of that loss, but it cannot fully account for the harm of being taken out of the market. Even were Lupin able to satisfy a judgment,⁹ "a defendant's ability to pay a judgment does not defeat a claim that an award of damages would be an inadequate remedy. Rather, a defendant's ability to pay merely indicates that a court should look to other considerations to determine whether a damages award will adequately compensate the patentee for the harm caused by continuing infringement." Apple III, 735 F.3d at 1369. And "[t]here is no reason to believe that [Lupin] will stop infringing, or that the irreparable harms resulting from its infringement will otherwise cease, absent an injunction." Robert Bosch LLC, 659 F.3d at 1155. The Court finds that remedies available at law would be inadequate.

9 Janssen argues that evidence in the record shows that Lupin's profits from generic sales of its darunavir products would be inadequate to satisfy a damages award for future infringement. See Janssen's 52(c) Br. at 12. Janssen has not submitted evidence concerning Lupin's other assets, which might or might not be sufficient to satisfy a judgment.

3. Balance of Hardships

The balance of hardships also weighs in favor of granting an injunction. If an injunction is not granted, Janssen will lose its Prezista business. Martori Tr. 2259:21–25; [redacted] The evidence clearly demonstrates that Lupin was aware of Janssen's '015 Patent, but chose to develop its infringing products anyway. See PTX 124, 125 (correspondence between Tibotec/Johnson & Johnson and Lupin disclosing the '015 Patent in response to Lupin's statement that it "propose[d] to start the commercial production of Darunavir ethanolate bulk drug in near future"). The Federal Circuit has made clear that "One who elects to build a business on a product found to infringe cannot be heard to complain if an injunction against continuing infringement destroys the business so elected." Robert Bosch LLC, 659 F.3d at 1156 (quoting Windsurfing Int'l, Inc. v. AMF, Inc., 782 F.2d 995, 1003 n. 12 (Fed.Cir.1986) (quotation marks omitted)). The Court concludes the balance of hardships weighs in favor of Janssen and granting an injunction.

4. Public Interest

The Court finds that "the public interest would not be disserved by a permanent injunction." eBay, 547 U.S. at 391, 126 S.Ct. 1837. The Federal Circuit "has long acknowledged the importance of the patent system in encouraging innovation. Indeed, the ‘encouragement of investment-based risk is the fundamental purpose of the patent grant, and is based directly on the right to exclude.’ " SanofiSynthelabo v. Apotex, Inc., 470 F.3d 1368, 1383 (Fed.Cir.2006) (quoting Patlex Corp. v. Mossinghoff, 758 F.2d 594, 599 (Fed.Cir.1985) ).

Lupin argues that the public interest will be harmed by precluding generic drugs at lower prices. See Joint Proposed Findings of Fact ¶ D1154 (quoting In re Barr Labs., Inc., 930 F.2d 72, 76 (D.C.Cir.1991) for the proposition that the Hatch–Waxman Act was designed to "get generic drugs into the hands of patients at reasonable prices-fast"). But the Federal Circuit recognizes that "while the statutory framework ... does seek to make low cost generic drugs available to the public, it does not do so by entirely eliminating the exclusionary rights conveyed by pharmaceutical patents. Nor does the statutory framework encourage or excuse infringement of valid pharmaceutical patents." Pfizer, Inc. v. Teva Pharms., USA, Inc., 429 F.3d 1364, 1382 (Fed.Cir.2005) ; see also Abbott Labs. v. Sandoz, Inc., 544 F.3d 1341, 1363 (Fed.Cir.2008) ("The patent laws promote ... progress by offering a right of exclusion for a limited period as an incentive to inventors to risk the often enormous costs in terms of time, research, and development." (quoting Kewanee Oil Co. v. Bicron Corp., 416 U.S. 470, 480, 94 S.Ct. 1879, 40 L.Ed.2d 315 (1974) (quotation marks omitted))). The record in this case supports the reasoning behind this statutory framework. Janssen invested over $1 billion to bring Prezista to market, Stoffels Tr. 112:23–25. And Dr. Stoffels testified that "if everyone can copy it the year we bring it to market or a few years afterwards, then there is no way we can justify that type of investment." Stoffels Tr. 113:11–13.

The Court finds that the public interest will not be disserved by granting a permanent injunction.

* * * * * *

The Court concludes that Janssen has met its burden to establish its entitlement to injunctive relief, and grants a permanent injunction enjoining Lupin from commercially manufacturing, selling or offering for sale, using, or importing the generic darunavir tablets described in Lupin's AND A in the United States, or any darunavir product that includes a bis-THF component made by any colorable variation of the process used to make Lupin's ANDA products, until after the expiration date of the '015 Patent.

B. The '411 Patent

i. Declaratory Judgment that the '411 Patent is Not Invalid and is Infringed by Mylan

Before trial, this Court granted Janssen's motion for summary judgment that Mylan infringes the '411 Patent. Summ. J. Op. at 5–8. At trial, Mylan failed to prove by clear and convincing evidence that claim 13 of the '411 Patent is invalid as obvious under 35 U.S.C. § 103, see supra Section II(C)(i)(5), or invalid as non-enabled under 35 U.S.C. § 112, ¶ 1, see supra Section II(C)(ii).

As a consequence, Janssen is entitled to a declaratory judgment that claim 13 of the '411 Patent is not invalid and that importing, selling, offering to sell, or using the generic darunavir tablets described in Mylan's ANDA would constitute infringement of the '411 Patent under 35 U.S.C. § 271(g). See Pre–Trial Order at 15.

ii. Permanent Injunction Barring Mylan from Importing or Selling its ANDA Products Before the '411 Patent Expires

Janssen also seeks a permanent injunction enjoining Mylan from commercially manufacturing, selling or offering for sale, using, or importing the generic darunavir tablets described in Mylan's ANDA or made by a process that is a colorable variation of the process claimed in the '411 Patent, until after the expiration date of the '411 Patent. Pre–Trial Order at 15.

The Court finds that Janssen has satisfied the four-factor test reaffirmed in eBay, 547 U.S. at 391, 126 S.Ct. 1837, and is entitled to the permanent injunction it seeks.

1. Irreparable Injury

The Court finds that, as with Lupin's infringement of the '015 process patent, Mylan's infringement of the '411 process patent will cause Janssen irreparable injury. Mylan makes largely the same arguments against this conclusion that Lupin does, with small differences, but the result is the same.a. Irreparable Harm

If Mylan is not enjoined from bringing its darunavir products made by an infringing process to market, Janssen will suffer the same irreparable harms identified with respect to Lupin and the '015 Patent : lost sales, price erosion, adverse impact on research and development, and having to compete directly with a product made by infringing its process.

[redacted] The Court concludes that, [redacted] Janssen will suffer near total loss of sales due to entry of Mylan's product made by infringing the '411 Patent.

[redacted] The infringement will result in a significant loss of funding for research and development. Stoffels Tr. 115:22–116:1, 116:12–14. Finally, as with Lupin, there is no question that Mylan and Janssen will be direct competitors, causing Janssen the irreparable harm of being driven out of the market by a product made by infringing its patented process. See Douglas Dynamics, 717 F.3d at 1345 ; Presidio Components, 702 F.3d at 1363.

The Court concludes these findings, taken together, constitute irreparable harm.

b. Causal Nexus

As with Lupin's infringement of the '015 Patent, the Court finds there is a causal nexus between Mylan's infringement of the '411 Patent and the irreparable harms.¹⁰ Mylan's process for making its generic darunavir products infringes the '411 Patent. Summ. J. Op. at 5–8. The sale of those products would lead directly to the irreparable harm-lost sales [redacted] price erosion [redacted] adverse impact on research and development (see Stoffels Tr. 115:22–116:1), and having to compete directly with an infringing competitor. Without infringing the '411 Patent, Mylan would not be able to produce the generics proposed in its ANDA, and without an injunction, it would use Janssen's patented process to destroy Janssen's Prezista business.

10 In the context of this patent, too, the Court rejects Janssen's argument that Congress addressed the causal nexus issue in enacting 35 U.S.C. § 271(g). See supra Section III(A)(ii)(1)(b)(2).

Mylan makes essentially identical arguments to those raised by Lupin concerning causal nexus. Mylan argues that Janssen will "lose sales if Mylan comes on the market irrespective of whether Mylan's process for making the active pharmaceutical ingredient in Mylan Pharms' ANDA products did or did not use the '411 patent's process steps," and that consumers will not care that the '411 process is used when deciding to purchase Mylan's generic products. See Joint Proposed Findings of Fact ¶¶ D1176, 1183–86. The Court rejects these arguments for the same reasons it rejected Lupin's. See supra Section III(A)(ii)(1)(b).

Mylan makes one additional argument. It argues that because Janssen does not practice claim 13 of the '411 Patent, Janssen "cannot tie Prezista sales" to infringement of that claim. Joint Proposed Findings of Fact ¶ D1178. But this relies on the same premise the Court rejects—that it is reasonable to equate the process at issue here with the features at issue in the Apple cases. Whether Janssen practices the claim should not impact the irreparable injury analysis in any event. See Trebro Mfg., 748 F.3d at 1171 ("In multiple instances, this court has held that a party that does not practice the asserted patent may still receive an injunction when it sells a competing product."); Presidio Components, 702 F.3d at 1363 ("[e]ven without practicing the claimed invention, the patentee can suffer irreparable injury" where there is "[d]irect competition in the same market"); Briggs & Stratton Corp. v. Chongquing Rato Power Co., Ltd., No. 5:13–cv–0316 (LEK), 2013 WL 3972391, at *22 (N.D.N.Y. July 23, 2013) ("The Federal Circuit has made it abundantly clear, however, that a patentee can demonstrate irreparable harm even if the patentee does not practice the asserted claims.").

11 ¹¹ The Court again rejects the argument, raised in the '411 context also, that Janssen has not demonstrated which entity sells Prezista. See supra Section III(A)(ii)(1)(b)(2).

The Court finds that Janssen has shown irreparable harm that is causally related to Mylan's infringement of the '411 Patent.

2. Inadequacy of Legal Remedies

Consistent with its reasoning in the context of the '015 Patent, the Court finds that remedies available at law would be inadequate to compensate Janssen for Mylan's infringement of the '411 Patent. Mylan and Janssen both agree that the inquiry here is the same as it is in the '015 context. See Joint Proposed Findings of Fact ¶¶ D1194, P579. As such, the Court concludes that the harm Janssen would suffer from Mylan's infringement of the '411 Patent cannot be adequately remedied by monetary damages. See supra Section III(A)(ii)(2); Robert Bosch LLC, 659 F.3d at 1155.

3. Balance of Hardships

The Court finds that the balance of hardships weighs in favor of Janssen and of granting an injunction. If an injunction is not granted, Janssen will lose nearly all of its Prezista business. [redacted] The Court credits the testimony of Dr. Reider, in conjunction with the European version of the ' 411 Patent produced from Mylan's files, in determining that Mylan was aware of and copied Janssen's process. See Reider Tr. 2047:11–2048:6, 2049:19–25; PTX 143 at MATRIX(Daru) 011139 (showing underlined portion of European version of '411 Patent ). The balance of hardships clearly weighs in favor of Janssen. See supra Section III(A)(ii)(3); Robert Bosch LLC, 659 F.3d at 1156.

4. Public Interest

For the same reasons discussed in the context of the '015 Patent, the Court finds that with respect to the '411 Patent the public interest will not be disserved by granting a permanent injunction. See supra Section III(A)(ii)(4).

* * * * * *

The Court concludes that Janssen has met its burden to establish its entitlement to injunctive relief, and grants a permanent injunction enjoining Mylan from commercially manufacturing, selling or offering for sale, using, or importing the generic darunavir tablets described in Mylan's ANDA or made by a process that is a colorable variation of the process claimed in the '411 Patent, until after the expiration date of the '411 Patent.

C. The '645 Patent

i. Declaratory Judgment that the '645 Patent is Not Invalid and is Infringed by Lupin

Before trial, Lupin stipulated to infringement of claim 4 of the '645 Patent. ECF No. 375. At trial, Lupin failed to prove that claim 4 of the ' 645 Patent is invalid, either under 35 U.S.C. § 103, see supra Section II(D)(i)(5), or § 112, ¶ 1, see supra Section II(D)(ii).

As a consequence, Janssen is entitled to judgment that claim 4 of the '645 Patent is not invalid, that Lupin has infringed the '645 Patent under 35 U.S.C. § 271(e)(2)(A) and that the making, using, selling, offering to sell, or importing of the generic darunavir tablets described in Lupin's ANDA would constitute infringement of the '645 Patent under 35 U.S.C. §§ 271(a). ECF No. 375.ii. Permanent Injunction Barring Lupin from Importing or Selling its ANDA Products Before the '645 Patent Expires

Janssen also seeks a permanent injunction enjoining Lupin from commercially manufacturing, selling or offering for sale, using, or importing the generic darunavir tablets described in Lupin's ANDA, or any colorable variations of the same, until after the expiration date of the '645 Patent and its associated exclusivity. Pre–Trial Order at 17.

The Court finds that Janssen has satisfied the four-factor test for entitlement to a permanent injunction. See eBay, 547 U.S. at 391, 126 S.Ct. 1837.

1. Irreparable Injury

The Court finds that Lupin's infringement of the '645 Patent will cause Janssen irreparable injury as readily as Lupin's infringement of the '015 process patent. The Court takes the two aspects of this inquiry in turn.

a. Irreparable Harm

If Lupin is not enjoined from bringing its infringing darunavir products to market, Janssen will suffer the same irreparable harms identified with respect to the '015 Patent : lost sales, price erosion, adverse impact on research and development, and having to compete directly with an infringing product.

As this Court has already found, see supra Section III(A)(ii)(1)(a), Lupin's entry of a generic into the market will lead to the loss of Janssen's sales, Martori Tr. 2259:21–25, significant price erosion, [redacted] adverse impact on research and development, Stoffels Tr. 115:22–116:1, 116:12–14, and being driven out of the market by an infringing product. See Douglas Dynamics, 717 F.3d at 1345 ; Presidio Components, 702 F.3d at 1363.

The Court concludes these findings, taken together, constitute irreparable harm.

b. Causal Nexus

The Court concludes that there is a causal nexus between Lupin's infringement of the '645 Patent and the irreparable harm. Lupin's darunavir products infringe claim 4 of the '645 Patent. ECF No. 375 (Stipulation). The sale of those products would lead directly to the irreparable harm—lost sales (see Martori Tr. 2259:21–25; [redacted] price erosion [redacted] adverse impact on research and development (see Stoffels Tr. 115:22–116:1), and having to compete directly with an infringing competitor. Without infringing the '645 Patent, Lupin would not have generics to sell, and without an injunction, it would use infringing products to destroy Janssen's Prezista business.

Unlike with the two process patents, Lupin's arguments concerning causal nexus have more bite in the context of the '645 Patent. Lupin is correct to argue that the novel, non-prior art element of the '645 Patent is that it is darunavir in an ethanolate solvate form. Janssen's witness, Dr. Myerson, testified that it was the advent of the ethanolate solvate form of darunavir that made the '645 Patent non-obvious. Myerson Tr. 1875:3–9. And Dr. Zingman testified that doctors do not prescribe Prezista because of its ethanolate form. Zingman Tr. 889:25–890:2; see also Leffler Tr. 2297:3–16 (testifying that "the limited physician surveys I've seen never listed ethanol solvate as one of the factors that play any role in a physician's prescribing habits"). Lupin also reiterates its argument that because Janssen would lose sales to Mylan, which does not infringe the '645 Patent, as easily as to Lupin, Janssen's "damages from the infringing acts are zero." Joint Proposed Findings of Fact ¶ D1214. Lupin thus argues that the ethanolate solvate feature does not drive demand for the product, meaning there is no causal nexus under the reasoning of the Apple cases.

This argument has some appeal, but it yields upon closer scrutiny. As the Court has found, this case is unlike the Apple cases because it does not involve "complex, multi-featured" products, but rather "relatively simple products" in terms of the number of features. See Apple III, 735 F.3d at 1362. The causal nexus inquiry "may be more easily satisfied" here. Apple III, 735 F.3d at 1362. The Court finds the example provided by the Federal Circuit in Apple III concerning a laptop and laptop battery especially applicable. As in that example, even taking as a given that doctors and patients do not prescribe or purchase Prezista (or would not prescribe or purchase Lupin's products) because it is in an ethanolate solvate form, "it is indisputable that the ability to" deliver darunavir into the body "is one of the reasons people buy" Prezista (indeed, likely the only reason). Apple III, 735 F.3d at 1364. "Thus, if the first person to invent a" way to deliver darunavir into the body in a medicine "had obtained a patent covering" that invention, "then it would be reasonable to say that the patented invention was a driver of demand for" darunavir. Id. This example is satisfied in this case. Dr. Wigerinck testified that he and his team at Tibotec eventually succeeded in preparing an ethanolate solvate form of darunavir, that it was the form in which they brought Prezista to market, that he searched for a non-solvated form of darunavir to bring to market without success, and that Tibotec never found something better than the ethanolate solvate form still used in Prezista today. See Wigerinck Tr. 332:3–6, 333:3–16, 335:6–336:5, 371:21–372:17, 494:11–495:17. The Court finds that the ethanolate form of darunavir is akin to the laptop battery of the Circuit's example, and can be considered a driver of demand for the product.¹² The causal nexus requirement of the irreparable injury inquiry is satisfied.

12 In addition, Dr. Wigerinck testified that he is the inventor on a separate patent covering the hydrate form of darunavir, U.S. Patent No. 8,518,987 B2, Wigerinck Tr. 498:21–499:8, and that Tibotec originally chose to pursue the ethanolate form over the hydrate form because that form was, at the time, difficult or impossible to make, Wigerinck Tr. 509:18–510:9. See Apple III, 735 F.3d at 1364 (continuing laptop example and stating that "if a particular patented laptop battery lasts significantly longer than any other battery on the market, then the replacement of that battery with a noninfringing battery might make a laptop less desirable. In that case, it might be reasonable to conclude that the patented battery is a driver of consumer demand for the laptop").

The Court adds that Janssen's Prezista business would be substantially destroyed by an infringing competitor, see Martori Tr. 2259:21–25; [redacted] which supports the conclusion that the causal nexus requirement is met. See supra Section III(A)(ii)(1)(b)(2).

13 ¹³ The Court again rejects the argument that Janssen has not demonstrated which entity sells Prezista. See supra Section III(A)(ii)(1)(b)(2).
--------

2. Inadequacy of Legal Remedies

The Court's analysis of the adequacy of legal remedies is the same with regard to Lupin's infringement of the '645 Patent as to Lupin's infringement of the '015 Patent. The Court finds that legal remedies are an inadequate remedy for Lupin's infringement. See supra Section III(A)(ii)(2); Robert Bosch LLC, 659 F.3d at 1155.

3. Balance of Hardships

As with the '015 Patent, the Court concludes that the balance of hardships clearly weighs in favor of Janssen with respect to the '645 Patent. See supra Section III(A)(ii)(3); Robert Bosch LLC, 659 F.3d at 1156. In addition, the practical effect of this injunction does not impose any hardship on Lupin because, as explained later, the Court is directing the FDA not to approve Lupin's ANDA until after the expiration of the '645 Patent. See infra Section III(C)(iii).

4. Public Interest

The Court finds that the public interest will not be disserved by granting a permanent injunction enjoining Lupin from selling products that infringe the '645 Patent. See supra Section III(A)(ii)(4). This has no practical effect on the public, since the Court is also directing the FDA not to approve Lupin's ANDA until after the expiration of the '645 Patent. See infra Section III(C)(iii).

* * * * * *

The Court concludes that Janssen has met its burden to establish entitlement to injunctive relief, and grants a permanent injunction enjoining Lupin from commercially manufacturing, selling or offering for sale, using, or importing the generic darunavir tablets described in the Lupin ANDA, or any colorable variations of the same, until after the expiration date of the '645 Patent and its associated exclusivity.

iii. An Order under 35 U.S.C. § 271(e)(4)(A)

Janssen also is entitled to an order under 35 U.S.C. § 271(e)(4)(A) directing the FDA not to approve Lupin's ANDA until after the expiration of the '645 Patent and the associated period of pediatric exclusivity.

Section 271(e)(4)(A) provides that upon a finding of infringement in Hatch–Waxman cases "the court shall order the effective date of any approval ... to be a date which is not earlier than the date of the expiration of the patent which has been infringed." 35 U.S.C. § 271(e)(4)(A) (emphasis added). Under this provision, an order under section 271(e)(4)(A) is mandatory upon a finding of infringement of a valid patent.

As the Federal Circuit has explained, "[i]f the patent holder proves infringement of a valid patent resulting from the filing of an ANDA, section 271(e)(4) provides three remedies." In re Omeprazole Patent Litig., 536 F.3d 1361, 1367 (Fed.Cir.2008). "Subparagraphs (B) and (C) provide the typical remedies for infringement: injunctive relief and damages." Id.

Subparagraph (A) ... provides an additional type of relief after a finding of infringement under section 271(e)(2) by requiring the district court to "order the effective date of any approval of the drug ... involved in the infringement to be a date which is not earlier than the date of the expiration of the patent which has been infringed."

Id. (emphasis added) (quoting 35 U.S.C. § 271(e)(4)(A) ). See also In re Rosuvastatin Calcium Patent Litig., 703 F.3d 511, 515 (Fed.Cir.2012) (where infringement is proven and "[a] challenge to the patent fails, the ANDA cannot be approved until expiration of the patent" (emphasis added)).

In the face of this clear authority, Lupin's reliance on a single district court decision for the proposition that this form of relief is discretionary is way off mark. See Joint Proposed Findings of Fact ¶ D1202 (citing SmithKline Beecham Corp. v. Apotex Corp., 247 F.Supp.2d 1011, 1045, 1052 (N.D.Ill.2003) ). This is especially so given that the Federal Circuit declared the issue of relief "moot" on appeal after invalidating the patent. SmithKline Beecham Corp. v. Apotex Corp., 403 F.3d 1331, 1346 (Fed.Cir.2005).

Because Lupin's ANDA products infringe claim 4 of the '645 Patent, and that claim is valid, this Court will enter an order under 35 U.S.C. § 271(e)(4)(A) directing the FDA not to approve Lupin's ANDA until after the expiration of the '645 Patent and any associated period of pediatric exclusivity.

D. Janssen's Request for an Additional Notification Provision

Janssen also requests that the Court include an order requiring Defendants to give Janssen and the Court 180 days' advance notice if either of them believes it can import and sell generic versions of darunavir that do not infringe the relevant patent(s). See Joint Proposed Findings of Fact P592–97. Janssen states that this notice is required because without it Janssen could not tell whether the products sold would infringe the '015 or '411 process patents, in particular. Id. ¶¶ P594–95. Janssen claims that Janssen and the Court will not have any way of making this determination and that "the Court's contempt powers will be a hollow sanction if the Court and Janssen are denied the information necessary to determine whether defendants are in compliance with the Court's orders." Id. ¶ P596.

The Court finds it unnecessary to impose the notification requirement Janssen requests. Janssen advances one case to support this remedy, see TiVo Inc. v. EchoStar Corp., 646 F.3d 869, 879 (Fed.Cir.2011), but there the district court had imposed a notification requirement only after finding the defendant in contempt of the court's previous order. The Court will not presume that Lupin or Mylan will violate this Court's orders. The existing statutory regime should continue to adequately protect Janssen, just as it did in the events leading up to its successfully preserving its rights and obtaining relief in this case.

The Court denies Janssen's request for an additional notification remedy.

IV. Conclusion

For the reasons set forth, this Court enters judgment: (1) that Lupin infringes claim 1 of the '015 Patent and claim 4 of the '645 Patent ; (2) that Mylan infringes claim 13 of the '411 Patent ; (3) that the asserted claims are not invalid; and (4) that Janssen is awarded the injunctive, declaratory and regulatory relief that it requested. Defendants' motion for judgment on partial findings of no remedy, ECF No. 841, and Lupin's motion for judgment on partial findings of invalidity due to lack of enablement of claim 1 of the '015 Patent, ECF No. 843, are both denied. An appropriate order follows.