Opinion
Civil Action 18-11026 (MAS) (DEA) (Consolidated)
09-20-2021
AMGEN, INC., Plaintiff, v. SANDOZ INC., al., Defendants. 1st Dose 2na Dose Day 6
NOT FOR PUBLICATION
OPINION (Filed under Seal)
MICHAEL A. SHIPP UNITED STATES DISTRICT JUDGE
This matter arises from Hatch-Waxman Act patent infringement claims brought by Plaintiff Amgen, Inc. (“Amgen”) against Defendants Dr. Reddy's Laboratories, Inc. and Dr. Reddy's Laboratories, Ltd. (together, “DRL”), Sandoz Inc., and Zydus Pharmaceuticals (USA) Inc. (“Zydus”) (collectively, “Defendants”). The Court has subject matter jurisdiction under 28 U.S.C. §§ 1331 and 1338, as well as under 28 U.S.C. §§ 2201 and 2202. The Court conducted a nine-day bench trial in this matter from June 14, 2021, to June 25, 2021. (See ECF Nos. 479-87.) The parties submitted post-trial briefing, proposed findings of fact, and proposed conclusions of law on July 8, 2021. (ECF Nos. 467-71.) Closing arguments were heard on July 28, 2021. (ECF No. 492.)
The Court consolidated docket numbers 18-11026, 18-11267, 18-11269, and 19-18806 fortrial. (Stipulation & Order regarding Case Consolidation for Trial, ECF No. 424.) Docket number 18-11026 is the lead matter. (Id.)
Amgen has asserted five patents against Defendants: U.S. Patent No. 7, 427, 638 (the “'638 Patent”), U.S. Patent No. 8, 455, 536 (the “'536 Patent”), U.S. Patent No. 10, 092, 541 (the “'541 Patent”), U.S. Patent No. 7, 893, 101 (the “'101 Patent”), and U.S. Patent No. 8, 093, 283 (the “'283 Patent”) (collectively, the “patents-in-suit”). The asserted patents concern stereomerically pure apremilast, associated methods and dosages for using stereomerically pure apremilast to treat psoriasis, and several crystalline forms of apremilast. Defendants have sought approval from the U.S. Food and Drug Administration (“FDA”) to market generic versions of apremilast. Amgen alleges that Defendants' generics will infringe the patents-in-suit.
This Opinion constitutes the Court's findings of fact and conclusions of law under Federal Rule of Civil Procedure 52(a). In evaluating the witnesses appearing at trial, and after the Court had the opportunity to hear their testimony and observe their demeanor, the Court undertook an individualized credibility assessment of each witness. The Court's findings of fact are based on those observations and credibility determinations of the witnesses, as well as a thorough review of the evidence admitted at trial. For the reasons stated below, the Court holds as follows:
1. As to the '638 Patent, Defendants have not met their burden of proving by clear and convincing evidence that the '638 Patent is invalid. As Defendants stipulated to infringement of the '638 Patent, judgment of infringement will be entered in favor of Amgen on the '638 Patent.
2. As to the '536 Patent, Defendants have not met their burden of proving by clear and convincing evidence that the '536 Patent is invalid. As Defendants stipulated to infringement of the '536
Patent, judgment of infringement will be entered in favor of Amgen on the '536 Patent.
3. As to the '541 Patent, Defendants have met their burden of proving by clear and convincing evidence that the '541 Patent is invalid. As such, judgment of non-infringement will be entered in favor of Defendants on the '541 Patent.
4. As to the '101 Patent, Amgen has not met its burden of showing Zydus's infringement by a preponderance of the evidence. As such, judgment of non-infringement will be entered in favor of Zydus on the '101 Patent.
5. As to the '101 Patent, DRL and Sandoz have not met their burden of proving by clear and convincing evidence that the '101 Patent is invalid. As DRL and Sandoz stipulated to infringement of the '101 Patent, judgment of infringement will be entered in favor of Amgen on the '101 Patent.
6. As to the '283 Patent, Zydus has not met its burden of proving by clear and convincing evidence that the '283 Patent is invalid. As Zydus stipulated to infringement of the '283 Patent, judgment of infringement will be entered in favor of Amgen on the '283 Patent.
(Zydus Stipulation of Facts ¶¶ 14-15, Exhibit A2 of the Final Pretrial Order, ECF No. 422-1; DRL & Sandoz Stipulation of Facts ¶¶ 9-12, Exhibit A3 of the Final Pretrial Order, ECF No. 422-2.).
(Zydus Stipulation of Facts ¶¶ 14-15, Exhibit A2 of the Final Pretrial Order, ECF No. 422-1; DRL & Sandoz Stipulation of Facts ¶¶ 9-12, Exhibit A3 of the Final Pretrial Order, ECF No. 422-2.)
(DRL & Sandoz Stipulation of Facts ¶¶ 9-12.)
(Zydus Stipulation of Facts ¶¶ 14-15.)
I. BACKGROUND AND FINDINGS OF FACT
A. The Parties
Plaintiff Amgen is a corporation organized under the laws of Delaware, with a principal place of business at One Amgen Center Drive, Thousand Oaks, California 91320. (Stipulation of Facts ¶ 1, Final Pretrial Order, ECF No. 425.)
Defendant Dr. Reddy's Laboratories, Inc. is a corporation organized under the laws of New Jersey, with a place of business at 107 College Road East, Princeton, New Jersey 08540. (Id. ¶ 3.)
Defendant Dr. Reddy's Laboratories, Limited is a company organized and existing under the laws of India, with a place of business at 8-2-337, Road No. 3. Banjara Hills, Hyderabad, Telangana 500034, India. (Id.)
Defendant Sandoz is a corporation organized under the laws of Delaware, with a place of business at 100 College Road West, Princeton, New Jersey 08540. (Id. ¶ 7.)
Defendant Zydus is a corporation organized under the laws of New Jersey, with a principal place of business at 73 Route 31 North, Pennington, New Jersey 08534. (Id. ¶ 8.)
B. Plaque Psoriasis, Psoriatic Arthritis, and Behcet's Disease
Plaque psoriasis is a chronic inflammatory or immune-mediated disorder of the skin, characterized by red, scaly patches (called plaques) that form on the skin. (Trial Tr. 276:12-23, 278:22-25 (Alexis).) Plaque psoriasis affects approximately three percent of the U.S. population. (Trial Tr. 277:18-21 (Alexis).) The condition can cause intense itching and discomfort. (Trial Tr. 277:2-3 (Alexis).) When psoriasis patients scratch or peel off the plaques, the plaques can easily bleed. (Trial Tr. 277:4 (Alexis).) Plaques can spontaneously shed, causing patients to complain about flakes on their clothing or other surfaces. (Trial Tr. 276:12-277:1 (Alexis).) Because of the visibility of plaques and the flaking, patients can also have negative social and psychological impacts. (Trial Tr. 284:23-285:24 (Alexis).) When the plaques affect patients' hands or feet, the plaques can impact their mobility and ability to perform common tasks at home. (Trial Tr. 276:10277:14 (Alexis).) Additionally, psoriasis can affect patients' joints, resulting in stiffness, pain, and reduced mobility in the form of psoriatic arthritis, which in turn affects about one-third of psoriasis patients. (Trial Tr. 277:10-14 (Alexis).) Another complication caused by psoriasis is Behget's disease, an incurable illness that causes painful oral ulcers. (Trial Tr. 278:7-16, 278:22-25 (Alexis).)
C. Otezla®
Amgen holds approved New Drug Application No. 205437 for 10 mg, 20 mg, and 30 mg, oral apremilast tablets, which are sold in the United States under the trademark Otezla® (“Otezla”). (Stipulation of Facts ¶ 10, Final Pretrial Order.) Under the prescribing information approved as of June 2020, Otezla tablets are FDA-approved to treat adult patients with moderate to severe plaque psoriasis who are also candidates for certain types of therapy, adult patients with active psoriatic arthritis, and adult patients with oral ulcers associated with Behqet's Disease. (Id. ¶ 11.)
The active ingredient in Otezla is apremilast. (Id. ¶ 12.) Apremilast can be represented by the chemical name “(+)-2-[l-(3-ethoxy-4-methoxyphenyl)-2-methylsulfonylethyl]-4-acetylaminoisoindoline-1, 3-dione” and the chemical structure shown below:
(Image Omitted) (Id. ¶ 13.) Apremilast is what's known as a phosphodiesterase type 4 (“PDE4”) inhibitor. (Id. ¶ 14.) PDE4 is a part of “11 families of phosphodiesterase, ” including, for example, PDE3 and PDE7. (Trial Tr. 142:6-12 (Schafer).) Apremilast treats plaque psoriasis by entering cells in the body, selectively inhibiting PDE4, and blocking the production of inflammatory cytokines such as Tumor Necrosis Factor alpha or TNFa; the result is an anti-inflammatory effect. (Trial Tr. 135:22-25, 136:20-137:12, 140:24-141:3, 168:19-169:2 (Schafer).) At the time of trial, apremilast was the only PDE4 inhibitor approved to treat psoriasis. (Trial Tr. 169:3-9 (Schafer), 304:9-305:2 (Alexis).)
D. Apremilast's Discovery and Relationship to Thalidomide
In the 1990s, Celgene Corporation (“Celgene”), then a small pharmaceutical company, operated a drug discovery program for “selective cytokine inhibitory drugs” or SelCIDS. (Trial Tr. 135:12-14 (Schafer).) Celgene conducted the SelCID program as part of its strategy to use the chemical structure of thalidomide molecules as a template for potential anti-inflammatory drugs. (Trial Tr. 131:2-15 (Schafer).) By 1999, it was known that thalidomide inhibited TNFa. (JTX-691, JTX-694.) It was also generally “known that TNF-alpha inhibition could be a way to treat diseases like psoriasis.” (Trial Tr. 137:7-13 (Schafer).)
Celgene was later acquired by Bristol Myers Squibb. (Trial Tr. 169:10-14 (Schafer).)
Notwithstanding its place in Celgene's SelCID research program, thalidomide was better known for its tragic role in medical history. In the 1950s and 1960s, thalidomide had been approved in Europe and elsewhere to treat morning sickness in pregnant women. (Trial Tr. 101:20-24 (Davies).) Thalidomide was later found to have teratogenic effects in fetuses-meaning it led to “debilitating, severe birth defects” in children after being taken by pregnant women. (Trial Tr. 101:17-24 (Davies).) It is undisputed that by 1999, although “there were several theories about the mechanism of teratogenicity[, ] . . . essentially it was not known” how thalidomide caused birth defects. (Trial Tr. 132:5-10 (Schafer).) Furthermore, as late as 2004, it was not known “which part of the thalidomide molecule was responsible for teratogenicity.” (Trial Tr. 102:23-25 (Davies).)
It was understood, however, that thalidomide is a racemic compound. (Trial Tr. 103:4-11 (Davies).) It was also believed that thalidomide's teratogenic effect was caused by its S-enantiomer. (Trial Tr. 105:6-20 (Davies).) Although the S-enantiomer of thalidomide was generally considered to be the cause of teratogenicity, the R- and S-enantiomers of thalidomide readily interconvert in the body due to the structure of thalidomide, which has an acidic hydrogen at the chiral center. (Trial Tr. 108:2-4 (Davies).) Indeed, scientific literature from 2001 rejected the suggestion that “the administration of a single thalidomide enantiomer rather than the racemic mixture . . . would improve [thalidomide's] side effect profile” because “thalidomide rapidly undergoes racemisation under both in vitro and in vivo conditions[, ] making administration of a single isomer non-viable.” (JTX-663-4.) Because of the interconversion of the R- and S-enantiomers of thalidomide, it was very difficult to separate the teratogenic properties of thalidomide from its therapeutic properties. (Trial Tr. 105:12-20, 106:10-24 (Davies).)
A racemic compound is also called a “racemate.” (Stipulation of Facts ¶ 112.) A racemate is a compound that is composed of a 50:50 mixture of two enantiomers. (Trial Tr. 94:12-14 (Davies).) Enantiomers are mirror-image molecules connected in the same sequence with different three-dimensional arrangements. (Trial Tr. 578:20-579:4 (Gribble).) Enantiomers can be differentiated experimentally by exposure to polarized light because a pair of mirror image enantiomers will rotate polarized light in different directions. (Trial Tr. 93:9-15 (Davies).) A chemist may refer to the enantiomer that rotates polarized light clockwise as the plus (+) enantiomer, and the enantiomer that rotates polarized light counterclockwise as the minus (-) enantiomer. (Trial Tr. 93:16-19 (Davies).) Alternatively, under a second convention, chemists may name an enantiomer with the labels “R” or “S.” (Trial Tr. 93:21-25, 94:1-10 (Davies).) “A racemic compound is a mixture of equal parts of enantiomers that is an equimolar mixture of two enantiomers such that the optical rotation is zero.” (Stipulation of Facts ¶ 111.)
The evidence at trial established that Dr. Hon-Wah Man, a Celgene researcher and chemist, synthesized apremilast on October 21, 1999 as a part of the SelCID research program. (JTX-1481 (Dr. Man's Compound Submission form for apremilast, known internally at Celgene as “CC-10004”); JTX-21044 (Celgene-created a slide deck presenting the history of apremilast); Trial Tr. 136:8-15, 144:2-3 (Schafer), 1303:16-1304:5, 1306:1-10 (Davies).) As discussed below, apremilast is an S-enantiomer of a racemic mixture known internally at Celgene as CC-7085. (Trial Tr. 143:22-144:1 (Schafer).) Apremilast's chemical structure has “what is called a phthalimido group, which is a common part between apremilast and the two enantiomers of thalidomide.”(Trial Tr. 103:1-11 (Davies).) Unlike thalidomide however, the “apremilast chiral center is not acidic and thus not readily racemizable like the chiral center found in thalidomide.” (Trial Tr. 108:1-4, 107:21-25 (Davies) (“[T]he glutarimide ring, which is a key pharmacophore for thalidomide analogs is not present in apremilast.”).) Accordingly, despite its structural similarities to thalidomide, apremilast does not racemize inside humans like thalidomide and is “not contraindicated for women of childbearing potential.” (Trial Tr. 301:16-17 (Alexis).)
Experts at trial alternatively referred to this group as “the thalimido group.” (See, e.g., Trial Tr. 658:2-3 (Gribble) (referring to “the phthalimide ring [that] is shared by both groups of compounds”), 1314:5-10 (Davies) (referring to the same group as a “thalimido group” which “is also present in thalidomide”).)
A pharmacophore is a part of a molecule that is thought to be at least in part responsible for the effect that the molecule has on a receptor within the body and is at least partly responsible for the observed biological effect. (Trial Tr. 108:16-22 (Davies).)
E. Apremilast's Efficacy, Safety, Potency, and Tolerability
Dr. Peter H. Schafer was a Celgene researcher involved in the SelCID program and an inventor listed on several of the patents-in-suit. (Trial Tr. 129:8-9, 136:10-15 (Schafer).) At trial, Dr. Schafer explained that selectivity for PDE4 was important to Celgene's SelCID program because “if you hit other PDEs, like PDE3 or PDE7 or any of the others, you may have additional side effects.” (Trial Tr. 142:6-12 (Schafer).) The uncontradicted evidence at trial was that by 1999, it was uncommon for companies that were researching PDE4 inhibitors to also focus on TNFa inhibitors. (See, e.g., 141:19-142:2 (Schafer).) In addition to PDE4 selectivity, Dr. Schafer testified that the SelCID program was also searching for drugs with sufficient “potency, ” “other drug-like properties, like absorption, stability, ” “activity in vivo in an animal model, ” and “safety in the animals.” (Trial Tr. 142:19-22 (Schafer).) Related to safety, Celgene also sought to develop antiinflammatory drugs with sufficient “tolerability”-that is, drugs that avoided less serious side effects such as headaches or nausea. (Trial Tr. 143:1-7 (Schafer).)
Celgene's 1999 synthesis of apremilast was preceded by its 1998 synthesis of a compound identified internally by the company as “CC-7085.” (Trial Tr. 139:11-24.) CC-7085 was the compound later identified as “Example 12” in U.S. Patent No. 6, 020, 358 (the “'358 Patent”), which was filed with the Patent Office in 1998 by Celgene. (Trial Tr. 1370:20-24 (Davies).) “CC-7085 is a racemic [compound]... which would be an equimolar ratio of the R[-] and S-enantiomer with the S-enantiomer being what is now known as apremilast.” (Trial Tr. 520:23-521:3 (Muller Dep. Designation), 143:22-144:1 (Schafer) (testifying that CC-7085 is related to apremilast in that “apremilast is an S-enantiomer with a structure that is related to the structure of CC-7085, which is a racemic mixture”).)
Early pre-clinical testing of apremilast demonstrated that it had greater potential use as an anti-inflammatory drug over CC-7085. According to a November 1999 report, preclinical studies indicated apremilast had a lower IC50-a measure of inhibitory potency-in both the TNFa and PDE4 assays than CC-7085. (Trial Tr. 146:20-147:8 (Schafer).) Dr. Schafer testified that this meant “that the S-isomer of CC-7085, ” identified internally by Celgene as “CC-10004, which is apremilast, was more active” than CC-7085 and other tested compounds. (Trial Tr. 146:22-147:5; JTX-1062.)
Another presentation reported experiments on PDE4A4 ratios associated with apremilast. A PDE4A4 ratio was a predictor of apremilast “being better tolerated.” (Trial Tr. 233:2-3 (Knowles); JTX-2O814.) Through these experiments, Celgene learned that apremilast had a low PDE4A4 ratio compared to the PDE4 inhibitors that were then under development by other companies. “Apremilast had a PDE4A4 ratio that was very low, 0.15, and that was more than ten-fold lower than Ariflo which was a PDE4 inhibitor that had made it furthest along in clinical development up to that point in time.” (Trial Tr. 151:10-16 (Schafer); JTX-20814.) Similarly, rolipram, an older PDE4 inhibitor, had an even higher PDE4A4 ratio. (Trial Tr. 151:18-20 (Schafer).) In addition to performing better than other companies' PDE4 inhibitors' PDE4A4 ratios, apremilast also performed ten-fold better than the CC-7085 racemate. (Trial Tr. 153:1-10 (Schafer).) At trial, Professor Richard Knowles, Amgen's expert in biochemistry, pharmacology, drug discovery, and PDE4 inhibitors, credibly testified that the PDE4A4 experiments would have been unexpected because “[t]here was no data to provide . . . anyone ... to think that that might be the result.” (Trial Tr. 195:4-8, 197:5-17, 198:10-16, 237:10-15 (Knowles).)
Researchers believed that the ratio of binding to the low affinity versus high affinity isoforms could predict whether nausea and vomiting were side effects associated with apremilast. (Trial Tr. 150:17-151:3 (Schafer); JTX-20814.)
Early murine shock model studies for apremilast also confirmed that in mice, apremilast was more than twenty-fold more potent than CC-7085. (Trial Tr. 154:9-25; JTX-11418.) As Dr. Schafer explained, in those studies, it appeared that a given dose of apremilast could produce a “50 percent reduction of TNFcc production in the mouse.” (Trial Tr. 154:11-18 (Schafer).) At trial, Dr. Schafer recounted his and his colleague's surprise: “[w]e were, frankly, very surprised at how much better apremilast was than the racemate .... Normally, if a racemate is a 50/50 mixture of two enantiomers, you might expect a two-fold difference in potency, all things being equal.” (Trial Tr. 155:1-15.) As Dr. Knowles explained, for apremilast to achieve such a large reduction in TNFcc over the racemate “suggests that the other enantiomer is blocking the effect of apremilast in some way. So that's really important. It tells you that you shouldn't expect to get good optimal effects with apremilast in the form of a racemate. You will get much better effects as a single enantiomer.” (Trial Tr. 240:2-8.)
Celgene's study of apremilast's effects on ferrets also produced unexpected results. By 1999, researchers understood that nausea and vomiting were among the most common side effects associated with PDE4 inhibitors. (Trial Tr. 250:21-251:1 (Knowles); JTX 677.) Ferret studies helped researchers because the studies evaluated both a compound's anti-inflammatory effect in ferrets' lungs and whether the compound produced vomiting. (Trial Tr. 157:11 -23 (Schafer).) The ferret studies also helped researchers determine what's known as a “therapeutic index, ” which is a measure reflecting the balance between a drug's comparative efficacy and its safety and tolerability. (Trial Tr. 158:1-19 (Schafer), 207:4-25 (Knowles).) Celgene's 2001 ferret studies unexpectedly determined that apremilast had a better than thirty-fold difference from the therapeutic index of comparable PDE4 inhibitors like cilomilast. (Trial Tr. 156:22-160:3 (Schafer), 236:15-19 (Knowles).)
The parties do not dispute that as of 1999 or 2002, the foregoing data relating to apremilast's potency, selectivity, safety, tolerability, and drug-like properties were not publicly available outside of Celgene to persons of ordinary skill in the relevant arts. (Trial Tr. 1690:4-1691:10 (Knowles).)
F. Competing Psoriasis Treatments
The FDA approved Otezla as an oral treatment for psoriasis in 2014. (Trial Tr. 279:3-6 (Alexis).)
Evidence at trial established that other pre-Otezla treatments for moderate to severe psoriasis had several risks and compliance barriers for patients. Amgen's dermatology expert, Dr. Andrew Alexis, testified that older oral systemics such as methotrexate, acitretin, and cyclosporine require “ongoing monitoring in the form of blood work and in some cases taking blood- measuring blood pressure.” (Trial Tr. 289:18-21 (Alexis).) Methotrexate's label warns that patients treated with this drug “should be closely monitored for bone marrow, liver, lung, and kidney toxicities.” (Trial Tr. 293:20-21 (Alexis); PTX-4131.) Indeed, patients treated with methotrexate are recommended to have a liver biopsy after a certain cumulative dose. (Trial Tr. 291:2-5 (Alexis).) Dr. Alexis also testified that methotrexate is associated with lymphoma and pulmonary fibrosis. (Trial Tr. 294:7-12 (Alexis).) Additionally, both methotrexate and acitretin are associated with fetal death and birth defects. (Trial Tr. 290:1-3.) Indeed, evidence at trial established that methotrexate is contraindicated for pregnant women and is rarely appropriate for women of childbearing potential. (Trial Tr. 295:23-296:12 (Alexis); PTX-4131.)
Further, for acitretin, initial lab work is needed to establish the patient's baseline lipid profile and liver function. (Trial Tr. 1755:2-19 (Alexis).) Cyclosporine is associated with a risk of cancer and a risk of permanent kidney damage, the latter risk increasing if the patient is treated with cyclosporine for longer than one year. (Trial Tr. 290:5-8, 291:6-12 (Alexis).) Dr. Alexis also described other barriers to treatment for methotrexate and acitretin, including that patients receiving those medications must abstain from drinking alcohol. (Trial Tr. 302:14-18 (Alexis).) Collectively, these pre-existing oral therapies were also known to suppress the immune system such that they were contraindicated for patients with other underlying diseases such as HIV, Hepatitis B, and Hepatitis C. (Trial Tr. 289:15-290:1, 304:16-19 (Alexis).)
Topical treatments such as creams, ointments, and gels were also available to treat psoriasis before Otezla's 2014 approval. (Trial Tr. 287:12-21 (Alexis).) Topical treatments had their own adherence barriers, including patient difficulty in applying the topicals to the skin, and the tendency of the topicals to transfer to clothing and other surfaces once applied. (Trial Tr. 289:511 (Alexis).) Relatedly, topical treatments were often not suitable for widespread psoriasis outbreaks because it was impractical for patients to apply the topicals to large areas of the affected skin. (Trial Tr. 288:14-17, 289:5-9 (Alexis).) Additionally, topical treatments had no efficacy in addressing psoriatic arthritis. (Trial Tr. 288:18-21 (Alexis).) Furthermore, topical treatments were not effective for patients with moderate and severe psoriasis, and topical corticosteroids lost their efficacy overtime. (Trial Tr. 288:22-289:4 (Alexis).)
Moreover, as of 2002, biologies, “which are injectable or infused medications, ” were also available to treat psoriasis. (Trial Tr. 287:12-25 (Alexis).) The biologic Remicade “is administered via intravenous infusion at an infusion center, ” which can be a treatment barrier for patients because it is not something that can be done at home. (Trial Tr. 291:14-19 (Alexis).) Other biologies such as Humira, Enbrel, and Stelara were self-administered by patients through subcutaneous injection. (Trial Tr. 290:13-16, 315:11-17 (Alexis).) Whether delivered by self-injection or intravenous infusion, Dr. Alexis testified that for many of his patients, these methods are a “significant barrier” because they may have a “phobia to needles or just simply have negative perceptions or just general concerns about a drug that has to be administered in that form.” (Trial Tr. 300:2-15 (Alexis).) Dr. Alexis testified that at least ten percent of his moderate to severe plaque psoriasis patients could not be convinced to undergo injectable therapies. (Trial Tr. 302:24-303:13 (Alexis).) The trial record established that, when biologies first became available for the treatment of psoriasis, they were distributed in pre-filled syringes that required the patient to be manually injected with a needle. (Trial Tr. 324:17-24.) Eventually, however, psoriasis biologies were sold in “auto-injectors” that “resemble pens” and where the needle “is more concealed than the traditional syringe.” (Trial Tr. 324:25-325:8 (Alexis).) Defendants' dermatology expert, Dr. Elaine Gilmore, testified that these devices deploy a small needle relative to the size of the syringe, which patients can barely feel and do not have to watch go into their skin. (Trial Tr. 885:4-14 (Gilmore).) In this way, biologies administered through auto injectors can help combat the needle hesitancy described by Dr. Alexis. (Trial Tr. 885:13-15 (Gilmore).)
Still, biologies involved other barriers and risks for patients. For example, biologies risked “reactivation of [latent] tuberculosis, and malignancies including lymphoma, ” as well as the risk of “serious infection.” (Trial Tr. 291:20-25 (Alexis).) Biologies also required periodic blood test monitoring and refrigeration. (Trial Tr. 291:25-292:12 (Alexis).) Notably, to the contrary, Dr. Gilmore testified that she saw very little incidence of patients rejecting biologic psoriasis treatments because they viewed lab monitoring as inconvenient. (Trial Tr. 894:24-895:11 (Gilmore).) Dr. Gilmore explained that this was because lab monitoring is “not something that is done super frequently. If s certainly not a weekly thing. It's not necessarily a monthly thing. Or even sometimes it's once a year.” (Trial Tr. 895:3-7 (Gilmore).) Moreover, Dr. Gilmore testified that the risks from biologies identified by Dr. Alexis such as serious infection, reactivation of tuberculosis, and malignancies “are very low risks” that treating physicians are capable of managing. (Trial Tr. 894:10-23 (Gilmore).)
It appears that Otezla itself had its own treatment barriers and risks to patients. According to Dr. Gilmore, forty to fifty percent of her patients treated with Otezla declined to continue with therapy due to associated gastrointestinal side effects that include nausea, vomiting, and diarrhea. (Trial Tr. 891:6-10.) Dr. Gilmore also testified that Otezla is associated with a “risk of an increase of adverse reactions of depression associated with Otezla therapy.” (Trial Tr. 890:4-22 (citing JTX-110).) Moreover, as Dr. Gilmore noted, Otezla “is dosed twice daily, ” once in the morning and once at night, which can be a barrier to treatment adherence for patients. (Trial Tr. 897:1-7 (Gilmore).) By contrast, after loading dose periods, biologies such as Humira and Enbrel are ultimately dosed every other week or weekly respectively. (Trial Tr. 896:16-22.) According to Dr. Gilmore, Stelara is injected with a syringe in the physician's office once every three months. (Trial Tr. 885:16-23 (Gilmore).)
Whatever the relative barriers and risks patients faced in using biologies, the record establishes that several biologies were more efficacious than Otezla. The record establishes that clinical studies of Otezla found that the drug had a PASI-75 score of 33.1% and 28.8% after two separate sixteen-week studies. (JTX-11012.) Enbrel, Remicade, Humira, and Stelara, which were approved before Otezla, each had higher PASI scores than Otezla:
A “PASI” score is a metric commonly used in clinical trials to measure a dermatologic treatment's efficacy. (Trial Tr. 319:21-320:1 (Alexis).) “PASI” stands for “Psoriasis Area and Severity Index.” (Trial Tr. 319:10-11.) A PASI score “takes into account the size of the area involved, as well [as the] redness of the plaques, thickness of the plaques, and scaling.” (Trial Tr. 823:1-4 (Gilmore).) “The scale goes from 0 to 72.” (Trial Tr. 823:10-12 (Gilmore).) “[I]n general, mild scores are defined by PASI scores of less than 10; moderate psoriasis would be considered a PASI score of 10 to 50; and greater than 50 would be reflective of a severe psoriasis level of involvement.” (Trial Tr. 823:12-16 (Gilmore).) A “PASI-75” score in a clinical trial refers to what percentage of people who were treated with a certain drug in a study achieved a seventy-five percent or greater reduction in their baseline PASI score at a certain endpoint (e.g., twelve or sixteen weeks) when compared to the patient's PASI score at the beginning of the trial. (Trial Tr. 886:24-887:8 (Gilmore), Trial Tr. 319:9-13, 320:2-21 (Alexis).)
(Trial Tr. 327:13-328:5, 330:18-331:7 (Alexis), 887:21-888:20 (Gilmore); DTX-3724-8; JTX-11012.)
Drug
Approval (psoriasis)
PASI-75 (%)
80%
Remicade®
2006
75%
88%
Humira®
2008
71%
78%
Stelara®
2009
66%
G. Defendants' ANDAs
DRL, Sandoz, and Zydus have each submitted Abbreviated New Drug Applications (“ANDAs”) to the FDA under 21 U.S.C. § 355(j), ANDAs Nos. 211756, 211658, and 211859, respectively. (Stipulation of Facts ¶¶ 74, 90, 94.) These applications seek FDA approval to engage in the commercial manufacture, use, and/or sale of apremilast tablets, generic versions of Amgen's Otezla products. (Id.)
Defendants' ANDAs each assert that several of Amgen's apremilast-related patents are invalid, unenforceable, or will not be infringed by the manufacture, use, offer for sale and/or importation of their respective ANDA products. Each of DRL's, Sandoz's, and Zydus's ANDAs assert that Amgen's '638 Patent, '536 Patent, '541 Patent, and '101 Patent are invalid, unenforceable, or will not be infringed by the manufacture, use, offer for sale, sale, or importation of DRL's ANDA products. (Stipulation of Facts ¶¶ 76-77, 92-93, 96-97.)
II. LEGAL STANDARD
Amgen must prove infringement by a preponderance of the evidence. See Octane Fitness, LLC v. ICON Health & Fitness, Inc., 572 U.S. 545, 557-58 (2014) (citing Bene v. Jeantet, 129 U.S. 683, 688 (1889)); Integrated Tech. Corp. v. Rudolph Techs., Inc., 734 F.3d 1352, 1356 (Fed. Cir. 2013); Siemens Med. Sols. USA, Inc. v. Saint-Gobain Ceramics & Plastics, Inc., 637 F.3d 1269, 1279, 1283 (Fed. Cir. 2011); Mannesmann Demag Corp. v. Engineered Metal Prods. Co., 793 F.2d 1279, 1282 (Fed. Cir. 1986); Envirotech Corp. v. Al George, Inc., 730 F.2d 753, 758 (Fed. Cir. 1984); Curlin Med. Inc. v. Acta Med., LLC, 228 F.Supp.3d 355, 358 (D.N.J. 2017) (citing Tech. Licensing Corp. v. Videotek, Inc., 545 F.3d 1316, 1327 (Fed. Cir. 2008)). “Determination of infringement is a factual question.” Cross Med. Prods., Inc. v. Medtronic Sofamor Danek, Inc., 424 F.3d 1293, 1302 (Fed. Cir. 2005) (citation omitted).
“[A] patent is presumed valid, and this presumption exists at every stage of the litigation.” Sanofi-Synthelabo v. Apotex, Inc., 470 F.3d 1368, 1375 (Fed. Cir. 2006) (citation omitted). Defendants must prove invalidity by clear and convincing evidence. See 35 U.S.C. § 282 (“The burden of establishing invalidity of a patent or any claim thereof shall rest on the party asserting such invalidity.”); Microsoft Corp. v. i4i Ltd., 564 U.S. 91, 102 (2011) (citation omitted); Amgen Inc. v. F. Hoffman-La Roche Ltd, 580 F.3d 1340, 1362 (Fed. Cir. 2009) (quoting PharmaStem Therapeutics, Inc. v. Viacell, Inc., 491 F.3d 1342, 1360 (Fed. Cir. 2007)); Tech. Licensing Corp., 545 F.3d at 1327; Purdue Pharm. Prods. L.P. v. Actavis Elizabeth LLC, No. 12-5311, 2015 WL 5032650, at *10 (D.N.J. Aug. 25, 2015) (quoting Linear Tech Corp. v. Int'l Trade Comm'n, 566 F.3d 1049, 1066 (Fed. Cir. 2009)).
When considering invalidity under an obviousness theory, the Court considers the totality of the prior art. Bristol-Myers Squibb Co. v. Teva Pharm. USA, Inc., 752 F.3d 967, 973 (Fed. Cir. 2014). “Obviousness is a question of law based on underlying facts.” Cross Med. Prods., 424 F.3d at 1302 (quoting Grp. One Ltd. v. Hallmark Cards, Inc., 407 F.3d 1297, 1303 (Fed. Cir. 2005)). When making an obviousness determination, courts must consider secondary considerations or objective indicia of non-obviousness. See id. at 1321-22. “[A] patent composed of several elements is not proved obvious merely by demonstrating that each of its elements was, independently, known in the prior art.” KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398, 418 (2007); see also Senju Pharm.. Co. v. Lupin Ltd, 780 F.3d 1337, 1341 (Fed. Cir. 2015). “Thus, a defendant asserting obviousness in view of a combination of references has the burden to show by clear and convincing evidence that a person of ordinary skill in the [art] [(“POSA”)] ... had reason to combine the elements in the manner claimed.” Senju Pharm. Co., 780 F.3d at 1341 (citing KSR Int'l Co., 550 U.S. at 418-19). Additionally, “a defendant must also demonstrate that a [POSA] would have [had] a reasonable expectation of success in combining the elements.” Id. (citing PharmaStem. Therapeutics, Inc., 491 F.3d at 1360).
Furthermore, in making obviousness determinations, courts must avoid hindsight analysis. In re Cyclobenzaprine Hydrochloride Extended-Release Capsule Patent Litig., 676 F.3d 1063, 1070-71 (Fed. Cir. 2012) (holding that courts should reject “hindsight claims of obviousness” where “a defendant urges an obviousness finding by merely throwing metaphorical darts at a board in hopes of arriving at a successful result, but the prior art gave either no indication of which parameters were critical or no direction as to which of many possible choices is likely to be successful” (internal quotations and alteration omitted) (quoting In re Knbin, 561 F.3d 1351, 1359 (Fed. Cir. 2009))); see also Mintz v. Dietz & Watson, Inc., 679 F.3d 1372, 1377 (Fed. Cir. 2012); Ortho-McNeilPharm., Inc. v. Mylan Labs., Inc., 520 F.3d 1358, 1364 (Fed. Cir. 2008) (hindsight is “always inappropriate for an obviousness test”).
“A determination that a patent is invalid as being anticipated under 35 U.S.C. § 102 requires a finding that each and every limitation is found either expressly or inherently in a single prior art reference.” Sanofi-Synthelabo, 470 F.3d at 1375 (internal quotations and citation omitted). “Anticipation is a question of fact.” Cross Med. Prods., 424 F.3d at 1302 (citation omitted).
III. DISCUSSION AND CONCLUSIONS OF LAW
A. The '638 Patent
The '638 Patent is entitled “(+)-2-[l-(3-Ethoxy-4-Methoxyphenyl)-2-Methylsulfonylethyl]-4-Acetylaminoisoindoline-l, 3-Dione, and Methods of Synthesis and Compositions Thereof.” (Stipulation of Facts ¶ 18.) The '638 Patent issued on September 23, 2008, from a patent application filed on April 13, 2005. (Id. ¶¶ 19, 22.) Peter H. Schafer, George W. Muller, and Han-Wah Man, were among the named inventors of the '638 Patent. (Id. ¶ 20.)
In the present litigation, Amgen, as the current assignee of the '638 Patent, asserts claims 3 and 6 of the '638 Patent against Defendants alleging infringement of the '638 Patent. (Id. ¶¶ 21, 23.) Claim 3 of the '638 Patent recites as follows: “[t]he pharmaceutical composition of claim 2 wherein said pharmaceutical composition is suitable for oral administration to a patient.” (Id. ¶ 28.) Claim 6 of the '638 Patent recites: “[t]he pharmaceutical composition of claim 5 wherein the amount of stereomerically pure (+)-2-[l-(3-ethoxy-4-methoxyphenyl)-2-methylsulfonylethyl]-4-acetylaminoisoindoline-1, 3-dione is from 10 mg to 200 mg.” (Id. ¶ 31.) As to claims 3 and 6, the parties have stipulated and agreed to the following claim construction: “a composition that comprises one stereoisomer of a compound and is substantially free of other stereoisomers of that compound, wherein that one stereoisomer is (+)-2-[l-(3-ethoxy-4-methoxyphenyl)-2-methylsulfonylethyl]-4-acetylaminoisoindoline-1, 3-dione.” (Id. ¶ 73.)
Claim 2 of the '638 Patent recites as follows: “[t]he pharmaceutical composition of claim 1 wherein said pharmaceutical composition is suitable for parenteral, transdermal, mucosal, nasal, buccal, sublingual, or oral administration to a patient.” (Id. ¶ 27.) Claim 1 of the '638 Patent is an independent claim and recites as follows: “[a] pharmaceutical composition comprising stereomerically pure (+)-2-[l-(3-ethoxy-4-methoxyphenyl)-2-methylsulfonylethyl]-4-acetylaminoisoindoline-1, 3-dione, or a pharmaceutically acceptable salt, solvate, or hydrate, thereof; and a pharmaceutically acceptable carrier, excipient or diluent.” (Id. ¶ 26.)
Claim 5 of the '638 Patent recites as follows: “[t]he pharmaceutical composition of claim 4 wherein the amount of stereomerically pure (+)-2-[l-(3-ethoxy-4-methoxyphenyl)-2-methylsulfonylethyl]-4-acetylaminoisoindoline-1, 3-dione is from 5 mg to 500 mg.” (Id. ¶ 30.) Claim 4 of the '638 Patent recites: “[t]he pharmaceutical composition of claim 2 wherein the amount of stereomerically pure (+)-2-[l-(3-ethoxy-4-methoxyphenyl)-2-methylsulfonylethyl]-4-acetylaminoisoindoline-1, 3-dione is from 1 mg to 1000 mg.” (Id. ¶ 29.)
1. Obviousness and the '638 Patent
At trial, Defendants attempted to demonstrate that claims 3 and 6 of the '638 Patent are invalid as obvious. The Court finds that Defendants have failed to meet their burden of showing by clear and convincing evidence that the '638 claims are invalid under an obviousness theory. See UCB, Inc. v. Accord Healthcare, Inc. (“UCB IH), 890 F.3d 1313, 1323 (Fed. Cir. 2018) (holding that invalidity under a theory of obviousness must be established by clear and convincing evidence).
35 U.S.C. § 103 provides that a patent may be invalidated if its claims are obvious in light of the prior art. Under the pre-AIA version of Section 103,
The parties agree that the Court's '638 obviousness determination is controlled by the pre-Leahy-Smith America Invents Act (“AIA”) version of 35 U.S.C. § 103 “because the patents-at-issue have effective filing dates prior to the March 16, 2013 effective date of the AIA”. (Defs.' Proposed Conclusions of Law ¶ 121 n.l, ECF No. 468; Pl.'s Proposed Conclusions of Law ¶67.)
A patent may not be obtained ... if the differences between the subject matter sought to be patented and the prior art are such that the subject matter as a whole would have been obvious at the time the invention was made to a person having ordinary skill in the art to which said subject matter pertains.35 U.S.C. § 103(a).
The obviousness determination “is centered on four factual inquiries, known as the Graham factors: ‘(1) the scope and content of prior art, (2) differences between claims and prior art, (3) the level of ordinary skill in pertinent art, and (4) secondary considerations such as commercial success and satisfaction of a long-felt need.'” Mitsubishi Tanabe Pharma Corp. v. Sandoz, - F.Supp.3d -, 2021 WL 1845499, at *10 (D.N.J. Apr. 7, 2021) (Wolfson, C.J.) (quoting Procter & Gamble Co. v. Teva Pharms. USA, Inc., 566 F.3d 989, 994 (Fed. Cir. 2009)); In re Ethicon, Inc., 844 F.3d 1344, 1349 (Fed. Cir. 2017) (“[T]he obviousness inquiry must ‘guard against slipping into use of hindsight and . . . resist the temptation to read into the prior art the teachings of the invention at issue.'” (quoting Graham v. John Deere Co., 383 U.S. 1, 36 (1966))).
The Court notes Amgen's argument that the Federal Circuit's controlling case law requires that the lead compound analysis be performed “in a case involving a new chemical compound.” (Pl.'s Post-Trial Br. 20, ECF No. 469.) As Defendants rightly note, however, the Federal Circuit has not required a lead compound analysis in cases involving the separation of enantiomers from racemic mixtures when “in the inventing process, there was no lead compound.” UCBII, 890 F.3d at 1313 (“We are not aware of any authority holding that a lead compound analysis is or is not required in cases involving purifying mixtures. Aventis simply required proving that a person of ordinary skill in the art would have been motivated to purify a mixture of compounds based on some known desirable property.”); see also Janssen Biotech, Inc. v. Celltrion Healthcare Co., No. 17-1108, 2018 WL 10910845 (D. Mass. July 30, 2018) (“[T]he Federal Circuit also held that while it was permissible to apply the lead compound test in the circumstances of UCB, the district court was not required to do so.”). Nothing before the Court suggests that Celgene's work began with a lead compound as a starting point.
The Federal Circuit has provided the following guidance to district courts performing an obviousness analysis as relating to purified and racemic compounds:
Where a claimed composition is a purified form of a mixture that existed in the prior art and “if the prior art would provide a person of ordinary skill in the art with reason to believe [it had desirable properties], the purified compound is prima facie obvious over the mixture even without an explicit teaching that the ingredient should be concentrated or purified.”UCBII, 890 F.3d at 1323 (alteration in original) (quoting Aventis Pharma Deutschland GmbHv. Lupin, 499 F.3d 1293, 1301 (Fed. Cir. 2007)). “Such a purified compound is not always prima facie obvious over the mixture, ” however, if, “for example, it may not be known that the purified compound is present in or an active ingredient of the mixture, or the state of the art may be such that discovering how to perform the purification is an invention of patentable weight in itself.” Aventis, 499 F.3d at 1301. Furthermore, “in cases involving new chemical compounds, it remains necessary to identify some reason that would have led a chemist to modify a known compound in a particular manner to establish prima facie obviousness of a new claimed compound.” Takeda Chern. Indus., Ltd. v. Alphapharm Pty., 492 F.3d 1350, 1357 (Fed. Cir. 2007).
“[E]ven though a patentee never must submit evidence to support a conclusion by a judge or jury that a patent remains valid, once a challenger introduces evidence that might lead to a conclusion of invalidity”-a prima facie case-“the patentee ‘would be well advised to introduce evidence sufficient to rebut that of the challenger.'” Pfizer, Inc. v. Apotex, Inc., 480 F.3d 1348, 1360 (Fed. Cir. 2007) (quoting Orthokinetics, Inc. v. Safety Travel Chairs, Inc., 806 F.2d 1565, 1570 (Fed. Cir. 1986)). Such rebuttal evidence may include “a comparison of test data showing that the claimed composition[] possess[es] unexpectedly improved properties or properties that the prior art does not have.” In re Dillon, 919 F.2d 688, 692-93 (Fed. Cir. 1990). Additionally, obviousness can be rebutted by showing that “the prior art teaches away from the claimed invention.” In re Geisler, 116 F.3d 1465, 1471 (Fed. Cir. 1997); see also Cheese Sys., Inc. v. Tetra Pak Cheese & Powder Sys., Inc., 725 F.3d 1341, 1352 (Fed. Cir. 2013) (“[E]ven when all claim limitations are found in prior art references, the fact-finder must determine what the prior art teaches, whether prior art teaches away from the claimed invention, and whether there was motivation to combine teachings from separate references.”).
a. Graham Framework: Level of Ordinary Skill in the Art
“Whether prior art invalidates a patent claim as obvious is determined from the perspective of one of ordinary skill in the art.” Star Sci., Inc. v. R.J Reynolds Tobacco Co., 655 F.3d 1364, 1374 (Fed. Cir. 2011) (citing KRS Int'l Co. v. Teleflex Inc., 550 U.S. 398, 420 (2007) (“The question is not whether the combination was obvious to the patentee but whether the combination was obvious to a person with skill ordinary in the art.”). The parties submit competing definitions of a POSA for the purposes of determining the validity or invalidity of the '638 Patent. (Compare Pl.'s Proposed Findings of Fact ¶¶ 270-72, ECF No. 470, with Defs.' Proposed Findings of Fact ¶¶ 400-03, ECF No. 467.) Nevertheless, Amgen submits that its experts' obviousness opinions would not change even accepting Defendants' definition of a POSA. (Pl.'s Proposed Findings of Fact ¶¶ 271-72.) Accordingly, the Court will use Defendants' definition of a POSA when considering the validity of the '638 Patent. That definition construes a POSA as follows:
[A] scientist with a Ph.D. in a field such as chemistry, medicinal chemistry, solid-state chemistry, pharmaceutics, pharmacology or the like, with one or two years of experience in the research, development, or characterization of pharmaceutical compounds, including chiral synthesis and resolution of stereomeric compounds or analytical methods (e.g., X-ray crystallography) for characterizing solid materials, or with an M.S. or similar degree and an additional three or more years of experience in the research, development, or characterization of pharmaceutical compounds.(Defs.' Proposed Findings of Fact ¶ 400.)
b. Graham Framework: Scope and Content of the Prior Art
According to Defendants, the '638 Patent claims “would have been obvious to a POSA over (1) the '358 patent and WO '606 and the knowledge of a POSA and (2) the '358 patent and Takeuchi and the knowledge of a POSA.” (Defs.' Post-Trial Br. 28, ECF No. 467.)
The '358 Patent is not among the patents-in-suit in the matter now before the Court. Celgene is the assignee of the '358 Patent. (DTX-1741.) Drs. George W. Muller and Hon-Wah Man, who were co-inventers of the '638 Patent, were also inventors of the '358 Patent. (Id.) The '358 Patent preceded the '638 Patent and was filed on October 30, 1998, and issued on February 1, 2000. (Id.) The '358 Patent broadly asserts that “[t]he compounds of the present invention are useful in the inhibition of phosphodiesterases, particularly PDE III and PDE IV, and in the treatment of disease states mediated thereby.” (DTX-1744, at 4:28-32; see also Stipulation of Facts ¶ 128 (same).) As Amgen frequently noted at trial, the '358 Patent discloses Formula I compounds, which cover billions of compounds. (See, e.g, Trial Tr. 619:13-19 (Gribble), 1313:11-15 (Davies).) Nevertheless, as Defendants often noted in response, the '358 Patent identifies seventeen example compounds that fall within the scope of the Formula I compounds, all of which are racemates. (Trial Tr. 619:9-12, 662:25-663:5 (Gribble), 1315:1-17 (Davies).)
Example 12 is one of these seventeen exemplar compounds. (Trial Tr. 594:23-24 (Gribble), 1398:16-20 (Davies).) Example 12 of the '358 Patent is a racemic mixture that is 50% of the (+) isomer (which is apremilast) and 50% the (-) isomer. (Trial Tr. 591:18-23, 624:12-13 (Gribble), 1315:8-1316:3 (Davies).) The '358 Patent states that “[t]he compounds of Formula I possess a center of chirality and can exist as optical isomers. Both the racemates of these isomers and the individual isomers themselves, as well as diastereomers when there are two chiral centers, are within the scope of the present invention.” (DTX-1746, at 8:63-67; Trial Tr. 594:6-17 (Gribble), 1406:6-1408:8 (Davies).) Although the '358 Patent claimed apremilast as an isomer within the scope of its inventions, it did not note any therapeutic utility in apremilast specifically. Nor did the '358 Patent suggest that apremilast was the active ingredient in the racemic mixture that had therapeutic utility. The '358 Patent noted only that Example 12 was one of seventeen racemic mixtures that had therapeutic utility.
Furthermore, although the '358 Patent does assert that certain of the example compounds can be used in “oral dosage forms includ[ing] tablets, capsules, dragees, and similar shaped, compressed pharmaceutical forms containing from 1 to 100 mg of drug per unit dosage, ” (DTX-1747, at 9:22-24), the '358 Patent makes no particular claims as to the suitability of apremilast for oral administration, as opposed to administration via intravenous or rectal administrations (id.). Without considering apremilast, however, the '358 Patent noted that Formula I racemates “can be used as such or can be separated into their individual isomers mechanically as by chromatography using a chiral absorbent” and “so as to obtain either or both substantially free of the other, i.e., in a form having optical purity of >95%.” (Trial Tr. 594:6-22 (Gribble); DTX-1746-7, at 8:67-9:12; DTX-1749, at 14:34-55.) Furthermore, evidence at trial established that the Patent Office considered the '358 Patent during the prosecution of the '638 Patent and found that the latter was neither anticipated nor rendered obvious by the '358 Patent. (JTX-16271-73; Trial Tr. 679:1-13 (Gribble).)
Defendants assert that the WO 01/34606 (“WO '606”) patent is prior art to the '638 Patent that “reinforces the teachings of the '358 Patent.” (Defs.' Post-Trial Br. 30.) WO '606 was published on May 17, 2001. (DTX-1591.) Man and Muller were also named inventors of WO '606, and Celgene is listed as the applicant on its cover. (Id.) Like the '358 Patent, WO '606 discloses a class of compounds called Formula I. (DTX-15911.) WO '606 “teach[es] the Formula I compounds in the '606 are useful to inhibit PDE4” and that “you can separate the individual isomers and obtain them substantially free of the other enantiomer in a form having an optical purity of greater than 95 percent.” (Trial Tr. 605:23-606:4 (Gribble).) Similar to the 358 Patent, WO '606 teaches that Formula I compounds “can be typified by oral dosage forms that include tablets, capsules, dragees, and similar shaped, compressed pharmaceutical forms obtaining from 1 to 100 mg of drug per unit dosage.” (DTX-15925, at 24:1-3; Trial Tr. 605:23-606:10 (Gribble).) Also like the '358 Patent, WO '606 notes non-oral methods of administering Formula I compounds. (DTX-15925, at 24:4-8.) It is undisputed that “there is overlap between the compounds of WO '606 and the '358 Patent, and Example 12 is within the scope of Formulae I, ” as described in both WO '606 and the '358 Patent. (Pl.'s Post-Trial Br. 33, ECF No. 469.) It is also undisputed, however, that unlike the '358 Patent, WO '606 fails to explicitly mention Example 12. (See, e.g, Trial Tr. 1320:20-21 (Davies) (“Q: Is Example 12 one of the 60 or so example compounds in WO '606? A: It is not, no.”).)
As for Takeuchi, which Defendants rely on as a part of their theory of obviousness, that paper is a 1999 academic publication in Organic Letters. (See generally DTX-168.) It is undisputed that Takeuchi does not describe Patent '358's Example 12 or any other compound listed in the '358 Patent. (Trial Tr. 1322:18-20 (Davies), 669:13-17 (Gribble).) Additionally, unlike the '358 Patent, Takeuchi does not provide any information about PDE4 inhibition. Takeuchi does, however, teach preparation of 3-fluorothalidomide, “a nonracemizable isoteric analogue of thalidomide.” (DTX-168J (abstract).) Takeuchi discloses that “[i]n strategies designed to develop drugs that might have the beneficial effects of thalidomide without the teratogenic side effects, synthesis of nonracemizable analogues of thalidomide has attracted much attention.” (DTX-1681.) Takeuchi teaches how to separate enantiomers of a compound that has a phthalimide ring, just as disclosed in the '358 Patent and with a separation greater than 99% purity. (Trial Tr. 610:6-15, 612:4-11 (Gribble).) The paper teaches preparation of 3-fluorothalidomide where the fluorine replaces the unstable hydrogen molecule in thalidomide and further teaches that 3-fluorothalidomide does not racemize. (Trial Tr. 609:10-23 (Gribble).) Takeuchi also teaches separation of the R- and S-enantiomers of 3-fluorothalidomide using a commercial chiral column and eluting with ethanol as the solvent, and that each enantiomer was obtained with an optical purity of more than 99%. (Trial Tr. 609:12-610:4.)
c. Graham Framework: Differences between Claims and Prior Art
The Federal Circuit has held that “[i]n cases involving the patentability of a new chemical compound, prima facie obviousness under the third Graham factor generally turns on the structural similarities and differences between the claimed compound and the prior art compounds.” UCBII, 890 F.3d at 1328 (citation omitted). “[Structural similarity between claimed and prior art subject matter, proved by combining references or otherwise, where the prior art gives reason or motivation to make the claimed compositions, creates a prima facie case of obviousness.” Aventis, 499 F.3d at 1301. Indeed, as the Federal Circuit has explained,
[I]f it is known that some desirable property of a mixture derives in whole or in part from a particular one of its components, or if the prior art would provide a person of ordinary skill in the art with a reason to believe that this is so, the purified compound is prima facie obvious over the mixture even without an explicit teaching that the ingredient should be concentrated or purified.Id.
Here, the Court finds that Defendants have failed to meet their burden of establishing that the prior art gave the POSA reason or motivation to resolve the Example 12 racemate into its enantiomers. Notwithstanding some indication of Example 12's therapeutic benefits as described in the '358 Patent, Defendants have not presented the Court with sufficient evidence to conclude that a POSA would have had reason to believe that the desirable properties of Example 12 derived in whole or in part from the apremilast isomer. Nor have Defendants demonstrated that the POSA would have had a reasonable expectation of success in resolving Example 12 in either 1999 or 2002.
The evidence at trial confirms these findings. The evidence established that the '358 Patent included “absolutely no biological data” on any of its example compounds, including Example 12. (Trial Tr. 1683:8-23 (Knowles).) Relatedly, the trial record also established that as of 1999 or 2002, there was no publicly available data concerning PDE4 potency, safety, tolerability, selectivity, or drug-like properties on any of the enantiomers of the '358 Patent's example compounds. (Trial Tr. 1690:20-1691:6 (Knowles).) Specifically, although the '358 Patent broadly discloses that the example compounds are “useful in the inhibition of phosphodiesterases, particularly PDE III and PDE IV, and in the treatment of disease states mediated thereby, ” (DTX-1744, at 4:28-32), evidence at trial established that such a claim would not provide aPOSA with sufficient reason to expend significant resources exploring the potential therapeutic utility of apremilast (see Trial Tr. 1370:6-15 (Davies)). Nor would this disclosure provide the POSA with information or reason to necessarily believe that an enantiomer of Example 12 had any such characteristics. As Amgen's synthetic, organic, and medicinal chemistry expert Dr. Stephen Davies explained, “ [Anything is possible really, so the racemate may have the same biological properties as the two enantiomers. It may have different biological properties.” (Trial Tr. 96:17-21 (Davies).) Moreover, having a PDE4-inhibiting racemate in hand would not reliably inform a POSA of either the potency or toxicity of either or both of its enantiomers. (Trial Tr. 96:21-24 (Davies) (“[T]he potency of one enantiomer-the potency might reside in one enantiomer, the toxicity might reside in the other, the potency and the toxicity might reside in both or just one.”).)
In rebutting the uncertain therapeutic value of a racemate versus its enantiomers, the Court finds Amgen's expert Dr. Davies more persuasive than Defendants' expert. Defendants' chemistry expert, Dr. Gordon Gribble, asserted that whenever one creates racemic mixtures of “any type in any kind of situation, the motivation is to separate those two compounds because you're dealing with a mixture of two things. You want to see what each enantiomer in this case is and what the biological activity is for each enantiomer.” (Trial Tr. 604:18-605:1 (Gribble).) This, according to Dr. Gribble, is “inherent motivation. There's never motivation not to separate racemic mixtures.” (Trial Tr. 605:1-3 (Gribble).) To the contrary however, Dr. Davies rejected the view that there would have been inherent motivation to resolve the Example 12 racemate, or any of the other example compounds referenced in the '358 Patent:
The idea that a POSA would go off and make 17 racemates, resolve 17 compounds through-each of which might take a huge amount of time and effort when they have no information about the . . . racemate ... if they were working for me in academia in one of my companies, they wouldn't be working for me much longer. The amount of effort and time and resources they would have wasted doing that would be unacceptable.
[I]f your racemate has some bad effects that you can't get around, then maybe it's a good idea to resolve it, if you know some information about the racemate that is bad for you. But if the racemate's fine, you would just carry on with the racemate.(Trial Tr. 1366:18-1367:5, 1487:19-1488:2.)
Relatedly, Amgen argues that even though the '358 Patent discloses that apremilast can be isolated from Example 12 by techniques known in the art, it does not disclose how the POSA could have isolated the Example 12 enantiomers without undue experimentation. Amgen notes that the '358 Patent does not provide a specific method for making stereomerically pure apremilast. (Trial Tr. 630:4-7 (Gribble).)
The evidence before the Court establishes that resolving a racemic mixture is a difficult process based on trial-and-error experimentation. (See, e.g., Trial Tr. 1350:6-13 (with Dr. Davies testifying that making enantiomers is “unpredictable. It's-it's trial and error. You have to do a lot of experimentation. Unless you're very lucky, you have to do a lot of experimentation to make enantiomers.”); JTX-1818 (“Although an experienced investigator can develop a ‘feel' for the resolution of a racemic mixture, in practice working out the experimental conditions is still a process of trial and error.”); JTX-1818 (“Selection of a resolution system is still a matter of trial and error and some resolutions have defied solution even after trying many different resolving agents and solvents.”); JTX-183 10 (“As of now, an optical resolution carried out in the usual manner is an empirical process.”). To be sure, Dr. Davies explained that virtually his entire academic career has involved projects to make single enantiomers and that he was often unsuccessful in these efforts. (Trial Tr. 1349:14-19 (Davies).)
Further, in the battle of the experts regarding the process of separating enantiomers/the Court finds Dr. Davies more persuasive than Dr. Gribble. As a threshold matter, both Amgen's and Defendants' experts appear to agree that the formation of chiral salts was not a viable method for separating the enantiomers of Example 12, notwithstanding the '358 Patent's representation to the contrary. (Trial Tr. 1363:5-14 (Davies), 645:15-19 (Gribble) (“Q: And, in fact, Example 12 is not amenable to the use of a chiral acid salt to isolate its enantiomers from the racemate; is that right? A: It is not an acid or not amine, not a base. So one cannot make a conjugate acid-base salt with Example 12.”).) Instead, Dr. Gribble asserts that chiral chromatography is the best and most efficient method for separating enantiomers of a racemic mixture. (Trial Tr. 586:9-12 (Gribble).) Dr. Gribble testified that separation of enantiomers using chiral chromatography is routine, standard experimentation taught in sophomore organic chemistry. (Trial Tr. 582:18-584:12, 593:4-594:5.) Drs. Gribble and Davies disagreed as to whether using chiral chromatography to resolve Example 12 is a mechanical separation requiring no chemical modification (Trial Tr. 585:20-586:1 (Gribble)), or whether such a resolution would be a chemical modification (Trial Tr. 1337:17-23 (Davies)). In any event, the parties appear to agree that chiral chromatography takes advantage of the selective interaction of enantiomers with a chiral adsorbent in the column. (Trial Tr. 585:20-586:1 (Gribble), 1356:14-21 (Davies).)
The '638 Patent provides that apremilast “can be isolated from the racemic compound by techniques known in the art. Examples include, but are not limited to, the formation of chiral salts and the use of chiral or high performance liquid chromatography ‘HPLC' and the formation and crystallization of chiral salts.” (JTX-310, at 9:13-16.)
As Dr. Davies explained, however, to utilize chiral chromatography to resolve Example 12, the POSA would have needed to find an appropriate solvent system for the chiral column to separate the enantiomers, and there were many such systems from which a POSA might have chosen in either 1999 or 2002. (Trial Tr. 1359:8-24 (Davies).) In determining how likely a given system was to succeed in accomplishing the selection, Dr. Davies testified that “nothing is predictable. You have to try. Trial and error.” (Trial Tr. 1355:10-13.) The POSA would have needed to consider the temperature, pH of the solvent system, and flow rate of the mobile phase in developing a chiral chromatography route. (Trial Tr. 1360:2-11 (Davies).) The conditions as to each of these variables would also have needed to be determined through experimentation. (Trial Tr. 1364:13-20 (Davies).)
Moreover, notwithstanding Takeuchi and its teaching on the ability to separate two enantiomers with a phthalimide ring with 99 percent purity using ethanol as a solvent, (Trial Tr. 610:5-15), Dr. Davies testified that he performed a literature search to see if he could find any reference or guidance to either the resolution of Example 12 and apremilast, or the resolution of any structural analogs via chiral chromatography; Dr. Davies found no such literature. (Trial Tr. 1363:19-23.) Whatever the teachings of Takeuchi, the Court cannot find that Defendants have met their burden of explaining how Takeuchi would have made the resolution of Example 12 via chiral chromatography obvious.
Significantly, in resolving the disputes between Dr. Davies and Dr. Gribble, the Court finds that, although both experts were credible, Dr. Davies's testimony was more persuasive. Dr. Gribble has had a long history of researching, teaching, and publishing in the fields of organic and medicinal chemistry, including synthesizing pharmaceuticals that eventually received FDA approval. (Trial Tr. 569:7-574:9 (Gribble).) Yet much of Dr. Davies's long career has involved practical experience specifically focused on separating enantiomers. (Trial Tr. 1490:8-10 (Davies).) His experience included establishing a company in 1992 called Oxford Asymmetry with the goal of helping pharmaceutical companies prepare single enantiomers. (Trial Tr. 1490:11-23.) Dr. Davies's experience has involved several successful resolution attempts and failures. (See, e.g, Trial Tr. 1349:14-1350:5 (“Well, there's one compound that I've been trying to resolve for at least two decades, and-and people . . . who join my research group, I get them to have a go .. . try a technique they want to try, whatever. And we have-we never succeeded in doing that.”).)
The matter before the Court is distinguishable from Aventis. In Aventis, the Federal Circuit considered whether a stereoisomer, in a form substantially free of other isomers, was obvious over the prior art disclosing its racemic mixture. 499 F.3d at 1301-03. The Federal Circuit ultimately concluded that the patent was invalid as obvious because at the time the racemic mixture was synthesized, it was understood that the stereoisomer was the racemic mixture's therapeutically active ingredient. Id. at 1302. Unlike in the present case, the prior art va Aventis provided the POSA with sufficient reasons to investigate the claimed stereoisomer because the prior art taught that there were substantial differences in its potency compared to the other stereoisomer. Id. (“[T]he Merck article taught that the SSS configuration of enalapril is 700 times as potent as the SSR form. The close structural analogy between 5(S) and SSSSR ramipril and SSS and SSR enalapril would have led a person of ordinary skill to expect 5(S) and SSSSR ramipril to differ similarly in potency.”).
The case here is more like Forest Laboratories, Inc. v. Ivax Pharmaceuticals, Inc. (“Forest Labs I”), 438 F.Supp.2d 479, 492 (D. Del. 2006), aff'd, 501 F.3d 1263, 1269 (Fed. Cir. 2007). In that case, the patent in dispute covered a substantially pure (+) enantiomer of citalopram, which was the active ingredient in an antidepressant. Id. The defendants contended that the patent was invalid as prima facie obvious based on an earlier patent that claimed racemic citalopram and disclosed its use as an antidepressant. Id. The defendants contended “that (+)-citalopram is - structurally similar to racemic citalopram and is used to treat the same condition as racemic citalopram, and therefore, one skilled in the art would have been motivated to make the claimed compound.” Id.
The district court rejected that argument. It noted that “even [defendants' experts acknowledged that the resolution of racemic compounds is difficult to achieve in today's world, and the activity of a particular enantiomer could not be known until it was actually separated and tested.” Id. at 493. The Court had also heard evidence of the “unpredictable nature of the separation techniques and separation results of racemates in general.” Id. Considering these challenges, the district court concluded that defendants had “not demonstrated by clear and convincing evidence that one skilled in the art would have been motivated to resolve citalopram.” Id. Moreover, the district court concluded that “a person skilled in the art seeking such a resolution would not have a reasonable expectation of success without undue experimentation.” Id. The Federal Circuit affirmed the district court's holding. Forest Lab 'ys, Inc. v. Ivax Pharms., Inc. (“Forest Labs II”), 501 F.3d 1263, 1269 (Fed. Cir. 2007) (“These findings fully support the conclusion that the claimed subject matter would not have been obvious to one of ordinary skill in the art.”).
Similarly, here, the Court finds that with limited publicly available data to a POSA in 1999 or 2002 on the biological effects of the '358 example compounds, the POSA would have had little reason or motivation to attempt to separate the compounds' enantiomers-let alone a reason to try to separate or synthesize the enantiomers of Example 12. That is especially so given the unpredictable “trial and error” nature of resolving racemic compounds. (Trial Tr. 1350:6-13.)The Court, accordingly, finds that this factor weighs against an obviousness finding in Defendants' favor.
The POSA also would have had limited motivation or reason to try and separate the enantiomers because a racemate may have either the same or different biological properties from the enantiomers. (See Trial Tr. 96:14-24; Sanofi-Synthelabo v. Apotex, Inc., 550 F.3d 1075, 1088 (Fed.Cir. 2008) (affirming district court finding that separation of enantiomers was not obvious because it “was not a simple or routine procedure and that success in separation, as well as the allocation of properties, was unpredictable”).)
d. Graham Framework: Objective Evidence of Non-Obviousness
A court's evaluation of the “secondary considerations” or “objective indicia of non-obviousness” “is not just a cumulative or confirmatory part of the obviousness calculus but [rather] constitutes independent evidence of non-obviousness.” Ortho-McNeil Pharm., Inc., 520 F.3d at 1365. Indeed, consideration of these criteria “help[s] inoculate the obviousness analysis against hindsight.” Mintz, 679 F.3d at 1378. Because the obviousness inquiry is “expansive and flexible, ” KSR Inf I Co., 550 U.S. at 415, there are a variety of secondary considerations that may be “utilized to give light to the circumstances surrounding the origin of the subject matter sought to be patented, ” including, but not limited to, “meeting a long-felt need, the inventors' success despite the failure of others, commercial success, copying, praise and recognition for the invention, unexpected results, and significant effort and serendipity.” Teva Neuroscience, Inc. v. Watson Labs., Inc., No. 10-5078, 2013 WL 12318005, at *7 (D.N.J. Sept. 20, 2013) (citing Ruiz v. A.B. Chance Co., 234 F.3d 654, 660-62 (Fed. Cir. 2000)). “For objective evidence of secondary considerations to be accorded substantial weight, its proponent must establish a nexus between the evidence and the merits of the claimed invention.” In re Huai-Hung Kao, 639 F.3d 1057, 1068 (Fed. Cir. 2011) (quoting Wyers v. Master Lock Co., 616 F.3d 1231, 1246 (Fed. Cir. 2010)). “[A] court need not find that all factors are present to determine that the objective considerations support a finding of non-obviousness.” Mitsubishi, 2021 WL 184599, at *21. “Although secondary considerations must be taken into account, they do not necessarily control the obviousness conclusion.” Apotex, Inc., 480 F.3d at 1372 (citing Newell Cos., Inc. v. Kenney Mfg. Co., 864 F.2d 757, 768 (Fed. Cir. 1988)).
i. Unexpected Results
“To be particularly probative, evidence of unexpected results must establish that there is a difference between the results obtained and those of the closest prior art, and that the difference would not have been expected by one of ordinary skill in the art at the time of the invention.” Bristol-Myers Squibb Co., 752 F.3d at 977. “While a marked superiority in an expected property may be enough in some circumstances to render a compound patentable, a mere difference in degree is insufficient.” Id. (citations and internal quotation omitted).
The parties disagree as to what is the closest prior art for the purposes of an unexpected results analysis. (See Pl.'s Post-Trial Br. 44 (“Defendants' argument that unexpected results can be compared only to data from the racemate, and not cilomilast, [is] baseless”); Defs.' Post-Trial Br. 50 (“Cilomilast was not the closest prior art, which would have been Example 12 of the '358 [P]atent.”).) The Court's review of UCB indicates that it need not limit its analysis of unexpected results to apremilast's improvements over the Example 12 racemate. UCB, Inc. v. Accord Healthcare, Inc. (“UCB I”), 201 F.Supp.3d 491 (D. Del. 2016) (analyzing an enantiomer claimed by the patent-in-suit for unexpected results over several other anti-epilepsy drugs, not just the enantiomer's racemate, which had been disclosed by prior art), aff'd, 890 F.3d 1313 (2018). Nevertheless, even accepting the premise that the Court's unexpected results analysis should be limited to considering evidence of apremilast's unexpected improvements over Example 12, evidence at trial supports finding in Amgen's favor. At trial, Dr. Schafer explained that in early murine shock model studies, it appeared that the chosen dose of apremilast could produce a “50 percent reduction of TNFa production in the mouse.” (Trial Tr. 154:11-18 (Schafer).) According to Dr. Schafer, “[w]e were, frankly, very surprised at how much better apremilast was than the racemate. We didn't expect a 20-fold difference in potency.. . . Normally, if a racemate is a 50/50 mixture of two enantiomers, you might expect a two-fold difference in potency, all things being equal.” (Trial Tr. 155:3-15; see Sanofl-Synthelabo, 492 F.Supp.2d at 390 (finding unexpected results where “the prior art did not enable a [POSA] to predict with a reasonable expectation of success whether one enantiomer. . . would have better pharmaceutical properties than the racemate itself, whether one enantiomer would have all of the activity and none of the toxicity of the racemate as a whole, ” or “whether a single enantiomer would have both all of the activity and all of the toxicity”).)
In all events, the trial record establishes that apremilast unexpectedly made substantial improvements over other PDE4 inhibitors in terms of efficacy and tolerability. A drug's “PDE4A4 ratio . . . was a predictor of being better tolerated” (Trial Tr. 233:1-2 (Knowles).) Another drug, Ariflo, also known as cilomilast, had a PDE4A4 ratio of 1.84 and rolipram had a PDE4A4 ratio of 3.64. (JTX-20814.) Apremilast's PDE4A4 ratio was tenfold better than Ariflo and twenty-fold better than rolipram. (Trial Tr. 152:17-18 (Schafer).) Regarding apremilast's therapeutic index for vomiting, Celgene's 2001 ferret studies unexpectedly determined that apremilast had a better than thirty-fold difference from the therapeutic index of cilomilast. (Trial Tr. 158:2-19 (Schafer), 236:15-19 (Knowles).) Based on the foregoing, the Court finds that apremilast's efficacy and tolerability over the Example 12 racemate and other PDE4 inhibitors was an unexpected success . . . especially taking into account the lack of prior art directing a [POSA] to that particular outcome.” Sanofi-Synthelabo, 492 F.Supp.2d at 392; cf. UCB I, 201 F.Supp.3d at 520-21, 543 (finding that unexpected results as to a drug's “increased therapeutic index” support a finding of unexpected results).
As for nexus, the Court agrees with Amgen that there is a nexus between the unexpected potency and side-effect profile of apremilast and claims 3 and 6 of the '638 Patent. In particular, the evidence before the Court establishes that apremilast's unexpected potency over the Example 12 racemic mixture derives from its separation from the other enantiomer (see Trial Tr. 155:3-15, 239:22-240:8), which is linked to the stipulated '638 Patent construction (see Stipulation of Facts ¶ 73 (“[A] composition that comprises one stereoisomer of a compound and is substantially free of other stereoisomers of that compound, wherein that one stereoisomer is (+)-2-[l-(3-ethoxy-4-methoxyphenyl)-2-methylsulfonylethyl]-4-acetylaminoisoindoline-l, 3-dione.”).
ii. Long-Felt, Unmet Needs
Next, with respect to long-felt but unmet needs, the record before the Court establishes that, before apremilast, there was a long-felt need for a psoriasis treatment that was suitable for oral administration to a patient and without the risks and barriers to adherence that were common with other psoriasis treatments. (See Trial Tr. 241:12-242:20 (Knowles).)
As discussed above, Amgen's dermatology expert, Dr. Alexis, testified that prior to 1999 or 2002 (the parties' competing proposed priority dates) and as late as Otezla's 2014 approval, topical treatments such as creams and ointments had several drawbacks for patients suffering from moderate to severe psoriasis, including topicals' messiness, patients' difficulty in applying them to large areas of affected skin, topicals' inability to treat psoriatic arthritis, and topicals' tendency to lose efficacy over time. (Trial Tr. 288:14-289:11 (Alexis).) Nevertheless, as Defendants argue, oral systemics such as methotrexate, acitretin, and cyclosporine predated apremilast. (Defs.' Post-Trial Br. 56-57.) And those drugs lacked the barriers to treatment present in topicals. (Id.)
But the older oral systemics had their own compliance barriers. The record establishes that older oral treatments like methotrexate had a similar efficacy in treating moderate to severe psoriasis as Otezla. (Trial Tr. 879:9-16 (Gilmore).) As Amgen demonstrated at trial, however, older oral systemics such as methotrexate, acitretin, and cyclosporine “require lab monitoring before and during and sometimes after” treatment. (Trial Tr. 290:21-291:1, 289:15-21 (Alexis).) Evidence at trial also established that older oral treatments like “methotrexate and cyclosporine . . . suppress the immune system” such that they were contraindicated for patients with other underlying diseases such as HIV, or those having underlying infectious conditions such as Hepatitis B, Hepatitis C, or latent tuberculosis. (Trial Tr. 289:15-290:1, 304:16-19 (Alexis).) Unlike these drugs, apremilast is not immunosuppressive and instead reduces inflammation, at least in part, through PDE4 inhibition. (Trial Tr. 169:3-9 (Schafer), 304:9-305:2 (Alexis).) The record at trial also establishes that older oral systemics such as methotrexate and acitretin are associated with fetal death and birth defects. (Trial Tr. 290:1-3 (Alexis).) In addition, Dr. Alexis testified that methotrexate is associated with lymphoma and pulmonary fibrosis. (Trial Tr. 294: 7-12 (Alexis).) Cyclosporine is also associated with a risk of cancer and permanent kidney damage, the latter risk increasing if the patient is treated with cyclosporine for longer than one year. (Trial Tr. 290:5-8, 291:6-12 (Alexis).) Similarly, patients treated with methotrexate are recommended to have a liver biopsy after a certain cumulative dose of the drug. (Trial Tr. 291:2-5 (Alexis).)
Moreover, the older oral systemics' risks and adherence barriers for psoriasis patients are documented in academic literature. As late as 2012, one academic wrote that apremilast's “lack of need for monitoring of liver and kidney function[] suggest that apremilast offers an efficacious treatment option with an acceptable safety and tolerability profile and improved convenience compared with present options and could help fill a gap in the management of psoriasis.” (DTX-1538.) To be sure, on cross-examination, Defendants' expert, Dr. Gilmore, agreed with the foregoing assessment that as late as 2012, apremilast possibly helped fill a gap in the treatment options for psoriasis. (Trial Tr. 944:7-18 (Gilmore).)
Nor do the newer biologies help show unmet needs. Newer biologies like Enbrel, Remicade, Humira, and Stelara, were FDA approved and effective for systemically treating psoriasis without the above risks and treatment barriers as early as 2004, before Otezla's 2014 FDA approval. (Trial Tr. 827:22-24, 828:10-11, 888:5-15 (Gilmore), 279:5-6 (Alexis).) It is undisputed, however, that the biologies had not begun to receive FDA approval for psoriasis before 1999 and 2002. (Trial Tr. 287:24-25, 299:10-14 (Alexis).) Defendants do not appear to argue, and cite no authority for the proposition, that the availability of biologic psoriasis treatments after the proposed priority dates for apremilast can nullify a demonstration of unmet need that existed prior to those priority dates. See Bristol-Myers Squibb Co. v. Teva Pharms. USA, Inc., 923 F.Supp.2d 602, 683 (D. Del. 2013) (“The Federal Circuit has recently explicitly stated that a court is to assess whether a long-felt and unmet need existed as of the ‘filing date of the challenged invention,' not as of ‘the time the invention becomes available on the market, when it can actually satisfy that need.'” (quotingProcter & Gamble Co., 566 F.3d at 998)).
Indeed, in Bristol-Myers, the district court considered whether a Hepatitis B drug, invented in 1990, entecavir, met a long-felt and unmet need. At the time of its invention there was no FDA-approved drug for the treatment of Hepatitis B. There were, however, “at least three other drugs . . . invented before entecavir that went on to gain FDA approval and be used with significant success to treat Hepatitis B (adefovir, tenofovir, and lamivudine).” Id. at 683. Notwithstanding these earlier drugs, the district court held that because it was not until 1991 that the first FDA approved drug for treating Hepatitis B came onto the market, “in this sense, there was a long-felt need for a drug that could be effective against Hepatitis B at [the] time entecavir was invented (a need that had not been met by any available drug that was actually in use).” Id. Similarly, here, the Court finds that because the biologies only became available for treating psoriasis after the invention of apremilast, the biologies do not nullify the long-felt and unmet need for an improved psoriasis medication at the time of apremilast's invention.
All in all, the Court agrees with Amgen that apremilast satisfied a long-felt and unmet need that existed in either 1999 or 2002. The Court also finds a nexus between the 638 Patent claims at issue in this matter and the unmet need. As Amgen argues, Otezla has met the need for an improved psoriasis treatment because of its active ingredient, stereomerically pure apremilast, as claimed in the '638 Patent. (See Trial Tr. 310:22-311:22 (Alexis).) Although the patentee generally bears the burden of establishing a nexus between satisfaction of long-felt, unmet needs and the merits of the patented invention, this nexus is presumed where, as here, “the patentee shows that the asserted objective evidence is tied to a specific product and that product ‘is the invention disclosed and claimed in the patent.'” WBIP, LLC v. Kohler Co., 829 F.3d 1317, 1329 (Fed. Cir. 2016) (quoting ZT Eaton & Co. v. Atl. Paste & Glue Co., 106 F.3d 1563, 1571 (Fed. Cir. 1997)); see also Brown & Williamson Tobacco Co. v. Phillip Morris Inc., 229 F.3d 1120, 1130 (Fed. Cir. 2000) (“[I]f the marketed product embodies the claimed features, and is coextensive with them, then a nexus is presumed and the burden shifts to the party asserting obviousness to present evidence to rebut the presumed nexus.”); (Trial Tr. 278:20-21 (Alexis) (testifying that the active ingredient of Otezla is apremilast); JTX-1108.)
iii. Others in the Field Tried and Failed to Develop PDE4 Inhibitors
Whether others failed to receive FDA approval for similar pharmaceutical compositions is also relevant evidence of a claimed pharmaceutical composition's non-obviousness. See, e.g., Knoll Pharm. Co., Inc. v. Teva Pharm. USA, Inc., 367 F.3d 1381, 1385 (Fed. Cir. 2004) (finding that the failure of two pharmaceutical companies to obtain FDA approval for similar pharmaceutical compositions was evidence of non-obviousness).
The record before the Court establishes that as of 1999 or 2002, “there were a lot of companies who were developing PDE4 inhibitors at that point in time.” (Trial Tr. 747:3-6 (Page).) Nevertheless, “a lot of them failed or didn't progress.” (Trial Tr. 747:17-18 (Page).) For example, as of 1999, atizoram, a PDE4 inhibitor developed by Pfizer, was discontinued because it was not effective at tolerable doses. (Trial Tr. 223:19-224:13, 225:4-8 (Knowles); JTX-6711; JTX-677.) Additionally, CD-840, a selective PDE4 inhibitor, “was discontinued in 1996 due to disappointing efficacy.” (Trial Tr. 224:14-20 (Knowles); JTX-677.) By 1999, more than one hundred PDE4 inhibitor candidates had been discussed in the published literature. (Trial Tr. 221:3-13 (Knowles).) But as of 1999, at least six PDE4 inhibitors had been discontinued. (Trial Tr. 214:3-19, 221:25-225:14 (Knowles).) And by March 2002, at least thirteen PDE4 inhibitors had been discontinued. (Trial Tr. 215:18-216:4, 221:25:14 (Knowles).) Ultimately, by the time of trial, at least 37 PDE4 inhibitors had been discontinued. (Trial Tr. 227:12-24 (Knowles).) Indeed, at the time of trial, “only two oral PDE4 inhibitors [had] been approved”: apremilast and roflumilast. (Trial Tr. 747:19-23 (Page).) The Court finds persuasive Dr. Knowles's opinion that if a PDE4 inhibitor was under development over a decade ago but has not yet been approved by the FDA, it is likely to have been discontinued. (Trial Tr. 227:17-21 (Knowles).) Moreover, the Court finds persuasive Dr. Knowles's testimony that even where it is not known why a drug was discontinued, it is unlikely that a drug “with good properties of efficacy, safety, drug-like properties, [etc.]” would be discontinued “for reasons unrelated to shortcomings with that compound.” (Trial Tr. 227:25-228:14.)
Notwithstanding the success of roflumilast, overall, the Court finds that the foregoing evidence establishes that the failure of other PDE4 inhibitors to receive FDA approval supports apremilast's non-obviousness as an orally administered pharmaceutical composition.
iv. Industry and Regulatory Skepticism
The Court also looks to “skepticism of skilled artisans before the invention.” Cheese Sys., Inc. v. Tetra Pak Cheese & Powder Sys. Inc., 725 F.3d 1341, 1352-53 (Fed. Cir. 2013). At trial, Amgen introduced evidence that the FDA and the pharmaceutical industry were skeptical about the safety of apremilast because of its relationship to thalidomide.
The uncontradicted evidence at trial established that the structures of apremilast and thalidomide are related to each other in that they both contain a phthalimido group. (Trial Tr. 103:2-11 (Davies).) As discussed above, thalidomide has a well-known history of teratogenic effects by causing birth defects when administered to pregnant women. (Trial Tr. 101:19-24 (Davies).) At trial, the parties' witnesses explained that as of 1999 or 2002, the mechanism of action for the teratogenicity of thalidomide was not known. (Trial Tr. 132:5-10 (Schafer), 658:23-25 (Gribble).) Nor was it known which part of the thalidomide molecule was responsible for teratogenicity. (Trial Tr. 102:23-25 (Davies).)
Amgen rightfully observes that the FDA's correspondence to Celgene about apremilast's structural relationship to thalidomide occurred in this context. In August 2004, in response to the Investigational New Drug Application Celgene was required to submit to the FDA before beginning a clinical trial for apremilast, the FDA asked Celgene to address several “safety concerns” at a then-upcoming teleconference. (PTX-8332.) These “safety concerns” included that “women of child-bearing potential” participating in the clinical study “should be using 2 acceptable forms of birth control.” (Id.) Additionally, the FDA asked Celgene to “comment regarding possible structural similarities of [apremilast] to thalidomide, ” “assess whether interconversion of the S-isomer [apremilast] to the R-isomer occurs in humans, ” and also comment on “[t]he fate of the metabolites of the parent drug” and “their potential for teratogenicity.” (PTX-8332-3; see also Trial Tr. 97:17-19 (“The FDA was skeptical about apremilast's structural relationship to thalidomide, raising concerns about teratogenicity and racemization.”).) The Court finds persuasive Dr. Davies's explanation that here, the FDA was “worried that apremilast, like thalidomide, may cause severe toxicity, particularly birth defects. And that apremilast, like thalidomide, may racemize in the body and that the resulting mixture-the racemic mixture, which would contain the opposite enantiomer may cause toxicity, particularly birth defects. And the toxicity may not be discovered until human studies.” (Trial Tr. 106:10-17 (Davies).) Celgene's minutes from the teleconference report that George Muller, one of the apremilast inventors, explained that apremilast's “chiral center was not acidic and thus not readily racemizable like the chiral center found in thalidomide”; nor had the data collected up to that point shown teratogenic effects. (JTX-2813.) Celgene's minutes further note that the “FDA seemed satisfied with Celgene's responses and said that they would convey them to the chemist, ” who, presumably prompted the FDA to raise these concerns. (Id.', Trial Tr. 108:23-109:13 (Davies).)
Notwithstanding the minutes' estimate that the FDA seemed satisfied with Celgene's responses, the Court finds that the fact that the FDA raised concerns about apremilast's connection to thalidomide is sufficient to establish skepticism by skilled artisans about apremilast before its invention. Cf. Bayer Intell. Prop. GmbH v. Aurobindo Pharm. Ltd., 2018 WL 3410020, at *14 (D. Del. July 13, 2018) (finding indicia of skepticism where, “[i]n particular, the European Medicines Agency (the European equivalent of the [FDA]) was skeptical” of a drug's safety profile due to its chemical structure and demanded that the patentee “demonstrate that there was no antibacterial effect. Chemists also expressed concern that the five-chlorothiophene moiety would produce toxic metabolites.”); Allergan, Inc. v. Sandoz, 726 F.3d 1286, 1291-92 (Fed. Cir. 2013) (holding that “FDA approval may be relevant to the obviousness inquiry” and “[t]he potential for FDA approval also may properly be considered ... in determining whether. . . there was skepticism regarding the efficacy of such a product”).
The Court also finds that evidence at trial established skepticism from the pharmaceutical industry toward apremilast. At trial, Dr. Schafer testified that between early 1999 and 2004, Celgene attempted to license its SelCID program, including apremilast, to other companies. (Trial Tr. 164:21-165:13 (Schafer).) Celgene sought to license its SelCID program because “Celgene was a small company” and it was “not able to really perform this drug development [of SelCIDS] on its own.” (Trial Tr. 165:4-5 (Schafer).)
GlaxoSmithKline (“GSK”) was one of several pharmaceutical companies that declined to license Celgene's SelCID compounds, and apremilast specifically. (Trial Tr. 156:22-157:10 (Schafer) (“We were hoping that [GSK] would be interested in the SelCID compounds and apremilast specifically and take a license to develop them or-or apremilast as a drug.”).) Amgen's biochemistry, pharmacology, and drug discovery expert, Dr. Richard Knowles, testified that as a GSK employee at the time, he was involved in evaluating the apremilast licensing opportunity presented to GSK by Celgene. (Trial Tr. 228:15-231:22 (Knowles).) Knowles recalled that in several of GSK's anti-inflammatory effect and tolerability studies, apremilast performed “outstandingly].” (Trial Tr. 230: 6-19 (Knowles).) Nevertheless, “[t]here was a mixture of views” in GSK about going forward with apremilast, with some in “the late development and commercial end of the organization in particular” believing “that the association with thalidomide was extremely unhelpful, ” potentially “giving rise to an actual risk that apremilast might also be teratogenic-like thalidomide.” (Trial Tr. 231:19-24 (Knowles).) Dr. Knowles further explained that some GSK officials were concerned about “the reputational risks associated with something that had a strong link to thalidomide and a structural element that was shared with thalidomide” and thought that licensing apremilast “was a significant risk to take.” (Trial Tr. 231: 24-232:6 (Knowles).) Ultimately, GSK declined to go forward with apremilast. (Trial Tr. 232:7-9 (Knowles).) Considering Dr. Knowles's testimony about GSK's consideration of apremilast licensing opportunities, the Court finds that there is evidence in the record of industry skepticism toward apremilast. Cf UCB 1, 201 F.Supp.3d at 517 (finding indicia of industry skepticism where, “[c]ompanies rejected [a class of compounds] because, ” among other things, “the compounds had not yet demonstrate[d] a lack of toxicity” (internal quotations omitted)).
v. Commercial Success
As Amgen's economic expert Dr. Chris Vellturo explained, there is little doubt that Otezla has achieved commercial success since receiving FDA approval. Roughly 1.7 million prescriptions were written for Otezla between the drug's 2014 launch and April 2020. (Trial Tr. 353.22-354.1 (Vellturo).) Within a year of its approval for psoriasis, Otezla had the highest share of prescriptions for systemic treatments among treatment-naive patients for psoriasis, i.e., patients who are either initiating systemic therapy for the first time or who are restarting therapy after an extended absence. (Trial Tr. 350:16-21, 356:10-15 (Vellturo).) In 2019, Otezla's prescription share among systemic psoriasis treatments surpassed methotrexate, even though methotrexate is available in generic form from a number of suppliers, and as a result has significant cost and insurance advantages over branded products like Otezla. (Trial Tr. 350:10-15 (Vellturo).)
Defendants do not appear to dispute that Otezla has been a commercial success. Rather, Defendants assert that Amgen cannot demonstrate a nexus between Otezla's marketplace performance and the patented attributes of stereomerically pure apremilast. (Defs.' Post-Trial Br. 60-63 (“Commercial success is relevant ‘only if there is proof that the sales were a direct result of the unique characteristics of the claimed invention-as opposed to other economic and commercial factors unrelated to the quality of the patented subject matter.'” (quoting In re Huang, 100 F.3d 135, 140 (Fed. Cir. 1996)).) Defendants assert that (1) “any evidence of Otezla's marketplace performance is not relevant to the non-obviousness inquiry because the '358 [P]atent was a blocking patent that precluded others from bringing a product like Otezla to the market” (Defs.' Post-Trial Br. 60); (2) “Amgen has not established a nexus between the commercial success and any novel aspects of the ['638] and ['536] Patents claims” (id. at 61); and (3) “Otezla's marketplace performance is also attributable to its marketing and promotional expenditures (id. at 62).
e. Summary of the Court's Obviousness Findings
Even assuming without deciding that Defendants' nexus arguments are meritorious with respect to commercial success, the Court would still find that the objective indicia weigh strongly in favor of non-obviousness. See Mitsubishi, 2021 WL 1845499, at *21 (“[A] court need not find that all factors are present to determine that the objective considerations support a finding of non-obviousness.”). With respect to other objective indicia in the record, including regulatory and industry skepticism toward apremilast because of its structural similarities to thalidomide, satisfaction of long-felt but unmet needs for oral psoriasis treatments, the failure of others to develop effective, safe, and tolerable PDE4 inhibitors that received FDA approval, and the unexpected success of apremilast over other PDE4 inhibitors, the Court finds a nexus between these indicia and the novel features of the '638 Patent. The Court also finds that although there is evidence that the prior art taught that thalidomide analogues could be useful in treating inflammatory diseases, the prior art also taught away from thalidomide analogues as a pharmaceutical composition based on its safety profile. Star Sei., Inc., 655 F.3dat 1374-75 (“[T]he fact-finder must not only determine what the prior art teaches, but whether prior art teaches away from the claimed invention and whether there is a motivation to combine teachings from separate references.” (citation omitted)).
After careful consideration of the Graham factors, the objective indicia, and the complete trial record, the Court finds that Defendants have failed to establish the obviousness of the '638 Patent by clear and convincing evidence.
2. Anticipation and the '638 Patent
Defendants' argument that the '638 Patent is invalid under a theory of anticipation also fails. “Claimed subject matter is ‘anticipated' when it is not new; that is, when it was previously known.” Sanofi-Synthelabo, 550 F.3d at 1082. “Invalidation on this ground requires that every element and limitation of the claim was previously described in a single prior art reference, either expressly or inherently, so as to place a person of ordinary skill in possession of the invention.” Id. “An anticipating reference must be enabling; that is, the description must be such that a person of ordinary skill in the field of the invention can practice the subject matter based on the reference, without undue experimentation.” Id. The Federal Circuit has held that “[t]he knowledge that enantiomers may be separated is not ‘anticipation' of a specific enantiomer that has not been separated, identified, and characterized.” See UCB II, 890 F.3d at 1330 (alteration in original) (emphasis added) (quoting Sanofi-Synthelabo, 550 F.3d at 1084).
The record before the Court establishes that the '358 Patent was filed with the Patent Office in 1998. (Trial Tr. 1319:16-18 (Davies).) It generally stated that the '358 Patent Formula I compounds:
possess a center of chirality and can exist as optical isomers. Both the racemates of these isomers and the individual isomers themselves, as well as diastereomers when there are two chiral centers, are within the scope of the present invention. The racemates can be used as such or can be separated into their individual isomers mechanically as by chromatography using a chiral absorbant. Alternatively, the individual isomers can be prepared in chiral form or separated chemically from a mixture by forming salts with a chiral acid, such as the individual enantiomers of 10-camphorsulfonic acid, camphoric acid, a-bromocamphoric acid, methoxy acetic acid, tartaric acid, diacetyltartaric acid, malic acid, pyrrolidone-5-carboxylic acid, and the like, and then freeing one or both of the resolved bases, optionally repeating the process, so as [to] obtain either or both substantially free of the other; i.e., in a form having an optical purity of >95%.(DTX-1746-7, at 8:63-9:12 (emphasis added).) Although this statement applied to the seventeen example compounds enumerated in the '358 Patent, it also applies generally to all compounds with a center of chirality, including the billions of compounds of Formula I. (Trial Tr. 620:1-22 (Gribble).) But the evidence at trial established that stereomerically pure apremilast had been synthesized no earlier than October 21, 1999. (JTX-21044; JTX-1481; Trial Tr. 144.2-3 (Schafer), 1303:5-1304:5 (Davies).) Accordingly, it appears that the above passage merely repeated the general knowledge that individual enantiomers can be separated from racemic mixtures.
Like the patent found valid in UCB II, the Court finds that the '358 Patent's disclosure of Example 12 “does not disclose its separation into individual enantiomers nor does it disclose any pharmaceutical data” of the apremilast enantiomer. UCB II, 890 F.3d at 1330. Because apremilast was not synthesized until a year after the filing of the '358 Patent, the prior art 358 Patent appears to report that Celgene prepared the Example 12 racemic compound “rather than the individual enantiomers.” Id. (emphasis omitted). The Court, accordingly, cannot find that the above passage from the '358 Patent anticipated the '638 Patent's claims as a matter of law.
Furthermore, the Court finds that Defendants have not demonstrated by clear and convincing evidence that the '358 Patent is enabled as to the separation of the Example 12 enantiomers. See Forest Labs II, 501 F.3d at 1268 (“A reference that is not enabling is not anticipating.”). Defendants assert that prior art references are presumed enabled. (Defs.' Post-Trial Br. 27 (citing Procter & Gamble Co. v. Nabisco Brands, Inc., 711 F.Supp. 759, 772 (D. Del. 1989)).) Nevertheless, “[a]ny presumption of enablement of prior art does not exclude consideration of whether undue experimentation would be required to achieve enablement. Sanofi-Synthelabo, 550 F.3d at 1085. The Federal Circuit has instructed district courts to look to the factors described in In re Wands for guidance as to whether undue experimentation would be required to achieve enablement. Id. (“The factors relevant to whether experimentation is undue are discussed in, e.g., In re Wands” and include “the quantity of experimentation that was actually needed, the amount of guidance provided in the reference, the presence or absence of actual examples of the experimental procedure, the state of the knowledge already available concerning the subject matter at issue, and the predictability or unpredictability in the specific area of science or technology”).
The Court finds that the Wands factors weigh in favor of a finding that undue experimentation was required to enable the '358 Patent. First, as to the quantity of expelimentation that was needed to make apremilast, Dr. Davies persuasively explained that “the time and cost of any necessary experimentation” in pursuing “[a]ny of the general routes listed in the '358 [P]atent would have required substantial time to make stereomerically pure apremilast.” (Trial Tr. 1387:5-9 (Davies).) Next, evidence at trial established that there was no guidance in the prior art as to a method for making stereomerically pure apremilast specifically. (Trial Tr. 1363:15-24 (Davies).) As for predictability or unpredictability, this factor weighs against a finding of enablement as the Court has already credited Dr. Davies's testimony that resolving a racemic mixture is a difficult process based on trial-and-error experimentation. (See, e.g, Trial Tr. 1350:6-13 (Davies).) Similarly, with respect to whether the '358 Patent discloses actual examples of the experimental procedure by which a POSA could make stereomerically pure apremilast, the Court finds such examples lacking. (See Trial Tr. 1387:23-1388:1 (Davies).)
Relatedly, as to the amount of guidance provided in the reference, the Court finds that the '358 Patent's guidance was minimal as to how to resolve Example 12. The '358 Patent states that the Formula I racemates “can be separated into their individual isomers mechanically as by chromatography using a chiral absorbant. Alternatively, the individual isomers can be prepared in chiral form or separated chemically from a mixture by forming salts with a chiral acid .... (DTX-1746-7, at 8:67-9:5).) The parties' competing experts appear to agree that the POSA would have known that the chiral acid salt separation route referenced in the '358 Patent would not work to separate the enantiomers of the racemate described in Example 12, because Example 12 is not a base. (Trial Tr. 1363:5-14, 1409:24-25 (Davies), 645:15-19 (Gribble).) As for chromatography or other methods, evidence at trial established that “[t]he '358 Patent does not provide any guidance in making stereomerically pure apremilast specifically.” (Trial Tr. 1388:3-5 (Davies).)
In finding that undue experimentation would have defeated enablement with respect to the '358 Patent as prior art, the Court's decision is consistent with Forest Labs I and Sanofi-Synthelabo. Defendants argue that “these two cases are factually distinguishable and legally inapposite” because in both cases, the asserted prior art references failed to disclose how to isolate or separate the isomers. (Defs.' Post-Trial Br. 25-26.) Here, Defendants assert that unlike the prior art at issue in Forest Labs I and Sanofi-Synthelabo, “the '358 [P]atent explicitly teaches that compounds of Formula I can be separated into individual isomers using chiral chromatography.” (Id. at 26.) But as frequently noted at trial, the Formula I compounds disclosed by the '358 Patent cover billions of compounds. (Trial Tr. 619:13-19 (Gribble), 1313:11-15 (Davies).) A statement that some indeterminate number of these compounds can be separated into individual isomers using chiral chromatography, without any additional guidance whatsoever as to the other variables such as temperature, pH of the solvent system, and flow rate of the mobile phase in developing a chiral chromatography route, (Trial Tr. 1360:2-11 (Davies)), is hardly instruction “such that a person of ordinary skill in the field of the invention can practice the subject matter based on the reference, without undue experimentation.” Sanofi-Synthelabo, 550 F.3d at 1082.
3. Priority Date of the '638 Patent
A patent is entitled to priority as of the date of conception of the invention, coupled with a reasonable diligence in reducing the invention to practice. Mahurkar v. C.R. Bard, Inc., 79 F.3d 1572, 1577 (Fed. Cir. 1996) (“[T]he person who first conceives, and in a mental sense, first invents . . . may date his patentable invention back to the time of its conception, if he connects the conception with its reduction to practice by reasonable diligence on his part, so that they are substantially one continuous act.” (second alteration in original) (internal quotations and citation omitted)). “To establish an actual reduction to practice . . . ‘the inventor must prove that (1) he constructed an embodiment or performed a process that met all the limitations of the claim; and (2) he determined that the invention would work for its intended purpose.” E.Idu Pont De Nemours & Co. v. Unifrax I LLC, 921 F.3d 1060, 1075 (Fed. Cir. 2019) (alterations omitted) (citation omitted). “[A]ctual reduction to practice, which constitutes in law the final phase of invention, cannot be established absent a showing of practical utility.” Fujikawa v. Wattanasin, 93 F.3d 1559, 1563 (Fed. Cir. 1996). “[W]hen testing is necessary to establish utility, there must be recognition and appreciation that the tests were successful for reduction to practice to occur.” Estee Lauder Inc. v. L'Oreal, S.A., 129 F.3d 588, 594-95 (Fed. Cir. 1997).
The Court finds that the composition claimed in the '638 Patent was first conceived on October 21, 1999. “Conception of a chemical substance requires knowledge of both the specific chemical structure of the compound and an operative method of making it.” Fina Oil & Chern. Co. v. Ewen, 123 F.3d 1466, 1473 (Fed. Cir. 1997). “Conception does not occur unless one has a mental picture of the structure of the chemical, or is able to define it by its method of preparation, its physical or chemical properties, or whatever characteristics sufficiently distinguish it.” Amgen, Inc. v. Chugai Pharm. Co., 927 F.2d 1200, 1206 (Fed. Cir. 1991). The evidence at trial established that Dr. Hon-Wah Man synthesized apremilast on October 21, 1999, at which time he had knowledge of the specific chemical structure of the chemical. (JTX-1481 (Dr. Man's Compound Submission Form for apremilast, known internally at Celgene as “CC-10004”); JTX-210 44 (Celgene-created slide deck presenting the history of apremilast); Trial Tr. 144:2-3 (Schafer), 1303:16-1304:5 (Davies).) The Court also notes that Dr. Man was “a member of a medicinal chemistry group whose goal was to synthesize compounds to be administered to . . . human beings, ” a fact which supports the Court's finding that by October 21, 1999, its inventors had conceived of apremilast as a pharmaceutical for human consumption. (Trial Tr. 1306:1-10 (Davies).) Furthermore, the fact that Celgene first successfully synthesized apremilast in October 1999, and then began testing apremilast “for enzymatic (PDE4) and cellular (TNF inhibition) activity” in November 1999, convincingly demonstrates that the company was in possession of an operative method for making the composition. (Trial Tr. 1306:1-10; JTX-21044; see, e.g., Merck Sharp & Dohme Corp. v. Actavis Lab 'ys FL, Lnc., No. 15-6075, 2017 WL 4366720, at *18 (D.N.J. Sept. 29, 2017) (“[T]here was substantial corroborating evidence, such as contemporaneous documentation and testimony, to show that the inventors of the' 151 patent had an ‘operable route' to synthesizing posaconazole.”).)
“Reduction to practice follows conception. To show actual reduction to practice, an inventor must demonstrate that the invention is suitable for its intended purpose.” Mahurkar, 79 F.3d at 1578 (citation omitted). “Depending on the character of the invention and the problem it solves, this showing may require test results.” Id. (citation omitted). Demonstrating reduction to practice also requires “the existence of sufficient evidence to corroborate inventor testimony regarding these events.” Medichem, S.A. v. Rolabo, S.L., F.3d 1157, 1169 (Fed. Cir. 2006). Courts have held that documents showing in vivo tests of drugs are adequate corroborative evidence of reduction to practice. See, e.g, Purdue Pharma L.P. v. Boehringer Ingelheim GMBH, 237 F.3d 1359, 1366 (Fed. Cir. 2001) (finding that evidence of “in vivo test use of the[] formulations in five randomized crossover bioavailability studies, ” among other evidence, “constitutes adequate proof of conception and reduction to practice”).
Here, Amgen has provided documentary evidence showing that the '638 claims had been reduced to practice by December 1999 following its October 1999 synthesis. Mahurkar, 79 F.3d at 1577 (“[T]he person who first conceives. . . may date his patentable invention back to the time of its conception, if he connects conception with its reduction to practice ... so that they are substantially one continuous act.” (citation omitted)). At trial, Dr. Schafer testified as to a December 1999 internal Celgene report documenting the results of a murine shock model study. (Trial Tr. 147:21-24, 148:9-14 (Schafer); JTX-1561.) Other internal reports from Celgene documented the results of the same murine shock model study, including the study's finding that apremilast could result in a “50 percent reduction of TNF-alpha production in the mouse.” (Trial Tr. 154:11-18 (Schafer); JTX-11418; JTX-11420).) This study showed apremilast was twenty-fold more potent than its racemate. (Trial Tr. 154:9-155:15 (Schafer).) Evidence at trial established that this result “implies that. . . there's enough compound there to have a sustained and potent anti-inflammatory effect. That may not translate one to one to human study, but it predicts a dose range which would be effective in human studies.” (Trial Tr. 239:9-19 (Knowles).)
Obviously, demonstrating apremilast's potency in mice is not the same as demonstrating its efficacy as a pharmaceutical composition suitable for oral administration to a patient as claimed in the '638 Patent. Nevertheless, evidence at trial established that the murine shock model was “the very first animal model that [Celgene] would test the compounds in, ” and was “part of the necessary regimen of testing to determine if a compound can be absorbed through the gut, if it's stable, if it reaches the target tissue in the cells, and if it has the desired effect.” (Trial Tr. 173:18-174:25 (Schafer).) Dr. Schafer credibly testified that if apremilast failed to work in that model, it would not go on for further testing in humans. (Trial Tr. 174:1-10 (Schafer).) Because animal studies such as the murine shock study described by Dr. Schafer are an essential part of reducing stereomerically pure apremilast to practice as a pharmaceutical composition suitable for patients within the dosage range claimed by claims 3 and 6 of the '638 Patent, and because this reduction occurred as early as December 1999, shortly after its October 21, 1999 synthesis, the Court is satisfied that Amgen is entitled to an October 21, 1999 priority date. Cf Cross v. Iizuka, 753 F.2d 1040, 1050-51 (Fed. Cir. 1985) (acknowledging “in vivo utility as sufficient to establish a practical utility, ” and explaining that “[t]his in vivo testing is but an intermediate link in a screening chain which may eventually lead to the use of the drug as a therapeutic agent in humans”).
4. Obviousness Type Double Patenting
The '283 Patent was filed on November 12, 2010 and issued on February 10, 2012. (Trial Tr. 1563:4-9 (Smith).) Defendants assert that claim 31 of the '283 Patent, like claims 3 and 6 of the '638 Patent, each state pharmaceutical compositions of stereomerically pure apremilast suitable for oral administration, with overlapping dosage ranges of 1 mg to 200 mg. (Trial Tr. 615:20-616:9 (Gribble); JTX-666-67, at 59:20-45, 61:1-3; JTX-321, at 31:27-39, 32:9-12.) At trial, Dr. Gribble testified that a POSA would “understand claim 31 to be within the scope of claims 3 and 6.” (Trial Tr. 616:14-16 (Gribble).) According to Dr. Gribble, the only difference between claim 31 of the '283 Patent and claims 3 and 6 of the '638 Patent is that claim 31 describes a crystalline form of apremilast whereas claims 3 and 6 do not specify the crystalline form. (Trial Tr. 616:10-13 (Gribble).) The FDA's Orange Book lists the expiration date of the '283 Patent as March 19, 2023. (Stipulation of Facts ¶ 58.) The '283 Patent claims priority to the application for U.S. Patent No. 6, 962, 940 (the “'940 Patent”), which was filed on March 19, 2003. (See PTX-1320.)
For its part, the '638 Patent issued on September 23, 2008, from a patent application filed on April 13, 2005. (Stipulation of Facts ¶¶ 19, 22.) Like the '283 Patent, the '638 Patent claims priority to the application for the '940 Patent. (See PTX-1320.) The FDA's Orange Book lists the expiration date of the '638 Patent as February 16, 2028. (Stipulation of Facts ¶ 25.) Although the '638 Patent issued earlier than the '283 Patent, the Orange Book lists it as expiring later than the '283 Patent. (See Defs.' Post-Trial Br. 38 (“Yet claim 31 of the '283 Patent expires on March 19, 2023, and claims 3 and 6 of the '638 patent expire on February 16, 2028.”).) Amgen argues that “the only reason the '638 Patent expires later than the '283 Patent is because” Amgen “received two statutorily authorized, and properly calculated, term extensions: a patent-term adjustment (or “PTA”) of 609 days, under 35 U.S.C. § 154(b), and a patent-term extension (or “PTE”) of 1, 186 days, under 35 U.S.C. § 156.” (Pl.'s Post-Trial Br. 52-53 (citations omitted).) At trial, Amgen presented evidence that each of these statutory time extensions were “properly awarded ... for delays outside Celgene's control.” (Id. at 56 (citing, among other things, Trial Tr. 1539:18-1540:12 (Smith).) It is undisputed that the two statutorily authorized time extensions that the '638 Patent received were properly calculated and awarded under applicable Patent Office policies, piactices, and procedures. (Trial Tr. 1555:2-9, 1558:21-1559:4 (Smith); PTX-1318; PTX-1319.)
Regardless of the reasons for which Amgen was awarded these adjustments, Zydus asserts that “Amgen's patents run afoul of the axiom that a patentee may not claim the same invention in different patents with different expiration dates under the judicially created doctrine of obviousness-type double patenting.” (Defs.' Post-Trial Br. 37.) Accordingly, Zydus argues that the patent-term adjustment awarded to Amgen under section 154(b) should be reduced by 609 days. (Id. at 39.)
Defendant Sandoz does not join this double-patenting defense. (Defs.' Post-Trial Br. 37 n.8.)
Notwithstanding these arguments, the Court finds that these statutorily authorized time extensions do not cause a patent to be invalid for obviousness-type double patenting (ODP). ODP is a “judicially created doctrine intended to prevent improper timewise extension of the patent right.” In re Braat, 937 F.2d 589, 592 (Fed. Cir. 1991) (emphasis omitted). A difference in expiration dates between two patents that arises solely from a statutorily authorized time extension, such as a patent-term adjustment pursuant to 35 U.S.C. § 154(b) or a patent-term extension pursuant to 35 U.S.C. § 156, cannot be the basis for an application of ODP. Novartis AG v. Ezra Ventures, LLC, 909 F.3d 1367, 1372-75 (Fed. Cir. 2018) (rejecting attempt to use “a judge-made doctrine” to “cut off a statutorily-authorized time extension”); Mitsubishi, 2021 WL 1845499, at *29-30. The '283 Patent is not a proper ODP reference for the '638 Patent because the difference in expiration dates between the '638 and '283 Patents arises solely from two statutorily authorized time extensions: a patent-term adjustment under 35 U.S.C. § 154(b) and another under 35 U.S.C. §156.
Even if the '283 Patent were a proper ODP reference for the '638 Patent, the Court would exercise its equitable discretion not to apply the doctrine of ODP under the circumstances of this case because the difference in expiration dates between the '638 and '283 Patents is not the result of prosecution gamesmanship or any improper conduct by Celgene. Immunex Corp. v. Sandoz, Inc., 964 F.3d 1049, 1059 (Fed. Cir. 2020) (noting that ODP is an “equitable doctrine”); Novartis Pharm. Corp. v. Breckenridge Pharm., Inc., 909 F.3d 1355, 1364 (Fed. Cir. 2018) (declining to apply ODP when difference in expiration dates was due to “happenstance of an intervening change in patent term law, ” rather than “prosecution gamesmanship” by the patentee); Gilead Seis., Inc. v. Natco Pharma Ltd., 753 F.3d 1208, 1210 (Fed. Cir. 2014) (applying ODP when patentee engaged in prosecution gamesmanship by structuring priority claims); Ln re Braat, 937 F.2d at 593.
B. The '536 Patent
Amgen asserts claim 6 of the '536 Patent against DRL, Sandoz, and Zydus. (Stipulation of Facts ¶ 37.) Amgen is the current assignee of the '536 Patent, which issued on June 4, 2013, from U.S. Patent Application No. 12/630, 788, filed on December 3, 2009. (Id. ¶¶ 35-36.) Claim 6 recites “[t]he method of claim 1, wherein the stereomerically pure compound comprises greater than about 97% by weight of (+) isomer based on the total weight percent of the compound.” (Id. ¶ 41 .)
Claim 1, in turn, recites “[a] method of treating psoriasis, which comprises orally administering to a patient having psoriasis about 10 mg to about 200 mg per day of stereomerically pure (+)-2-[l-(3-ethoxy-4-methoxyphenyl)-2-methylsulfonylethyl]-4-acetylaminoisoindoline-l, 3-dione, wherein the compound is administered in the form of a tablet or capsule as either a single dose or a divided dose.” (Id. ¶ 40.)
In response to Amgen's assertion, Defendants challenge the validity of the '536 Patent, arguing that “[t]he clear and convincing evidence presented at trial renders claim 6 of the '536 [P]atent invalid for anticipation and/or obviousness” as of its March 20, 2002 priority date for many of the same reasons discussed above with respect to the '638 Patent.” (Defs.' Post-Trial Br. 39.)
Amgen does not appear to contest Defendants' proposed priority date for the 536 Patent. (See Defs.' Proposed Findings of Fact ¶ 707 (noting that both Amgen and Defendants' experts used a priority date of March 20, 2002 for the '536 patent, which is the effective filing date of Provisional Application No. 60/366, 515. (Trial Tr. 699:5-11 (Page), 866:16-18 (Gilmore), 1746:5-8 (Alexis).)
1. Anticipation and the '536 Patent
For the reasons discussed above with respect to the '638 Patent, the Court finds that the '358 Patent does not anticipate Claim 6 of the '536 Patent. (See Defs.' Post-Trial Br. 39-41.) Defendants again point to general statements in the '358 Patent that the billions of compounds of Formula I, ” including the seventeen examples listed, “are used ... to inhibit the undesirable effects of TNFa and PDE IV” and can be “administered orally, rectally, or parenterally, alone or in combination with other therapeutic agents including antibiotics, steroids, etc., to a mammal in need of treatment.” (DTX-1745, at 7:1-6.) Additionally, Defendants direct the Court to statements in the '358 Patent that generally observe that “[d]ecreasing TNFa levels” and “inhibiting PDE IV thus constitutes valuable therapeutic strategies for the treatment of many inflammatory, infectious immunological or malignant diseases” including “psoriasis, ” among many others. (DTX-1745, at 4:35-49.) As discussed at length above, although apremilast's racemate is the Example 12 compound given in the '358 Patent, the '358 Patent “does not disclose its separation into individual enantiomers, nor does it disclose any pharmaceutical data” on the apremilast enantiomer's ability to treat psoriasis. UCBII, 890 F.3d at 1330. Once again, “[t]he knowledge that enantiomers may be separated is not ‘anticipation' of a specific enantiomer that has not been separated, identified, and characterized.” Id. at 1323 (alteration in original) (emphasis added) (citation omitted).
2. Obviousness and the '536 Patent
Regarding obviousness combinations, in some respects Defendants' proposed '536 Patent obviousness combinations overlap with the '638 Patent discussed above. Defendants, however, also assert additional prior art references. (Defs.' Post-Trial Br. 41 (asserting that the 358 Patent “not only anticipates the asserted claim of the '536 Patent, but also renders it obvious, as of the '536 patent's priority date, ” when “viewed in combination with the knowledge of a POSA and the teachings of: (1) WO '606, Dyke 1999, and Marriott 2001; (2) Takeuchi, Dyke 1999, and Marriott 2001; and/or (3) Dyke 1999, Marriott 2001, and Muller 1998”).)
Hazel Dyke & John Montana, The Therapeutic Potential of PDE4 Inhibitors (“Dyke 1999”) is a 1999 article that provides a comprehensive review of the therapeutic potential of selective PDE4 inhibitors in a variety of diseases. (Trial Tr. 710:21-711:8 (Page). See generally JTX-67.) At trial, Defendants' pharmacology and drug discovery expert, Dr. Clive Page, testified that Dyke 1999 gave psoriasis “its own section in this review, suggesting that people working in this industry at this time were considering this as a sensible disease area to use PDE4 inhibitors as a possible treatment.” (Trial Tr. 711:23-712:5 (Page).) Defendants note that Dyke 1999 asserts “the potential utility of PDE4 inhibitors in [rheumatoid arthritis] and the broad-spectrum anti-inflammatory action of such compounds suggests that they have the potential to provide a beneficial treatment for psoriasis.” (JTX-6712; Trial Tr. 712:8-18 (Page).) Dyke 1999 also disclosed an instance of PDE4 inhibitors being used to treat psoriasis. (JTX-6712.) In that instance, topical administration of “[t]he selective PDE4 inhibitor Ro20-1724, was investigated in two double-blind studies.” (Id.', Trial Tr. 712:20-713:4 (Page).) The studies showed that PDE4 inhibition “improve[d] psoriatic lesions” and “had no adverse systemic or cutaneous effects, suggesting the therapeutic potential of such compounds in the treatment of psoriasis.” (JTX-6712; Trial Tr. 712:20-713:4 (Page).)
J. Blake Marriott et al., Immunotherapeutic and Antitumor Potential of Thalidomide Analogues (“Marriott 2001”), is a 2001 article. (Trial Tr. 715:14-716:3. See generally JTX-66.) Marriott 2001 generally discloses that thalidomide was “established as an effective immunomodulatory and anti-inflammatory drug” that “show[ed] potential for the treatment of a range of conditions, including rheumatoid arthritis.” (JTX-662-3; Trial Tr. 716:15-20 (Page).) According to Marriott 2001, thalidomide had been shown to inhibit synthesis of TNFa, which is “a key regulator of other pro-inflammatory cytokines and leukocyte adhesion molecules and therefore represents a therapeutic target in a number of conditions where the overproduction of TNFa is associated with a pathological inflammatory cascade.” (JTX-663.) Marriott 2001 reported that “[i]n the last few years two anti-TNFa biological agents have received FDA approval. Enbrel, a TNFa receptor received approval for the treatment of [rheumatoid arthritis] and infliximab, a TNFa antibody received approval for Crohn's disease.” (Id.) Additionally, Marriott 2001 reported on a SelCID that had entered into clinical development and that this SelCID, a Celgene compound known as CDC-801, had “successfully completed two Phase I clinical trials in the UK, ” with no serious adverse events reported in the second trial. (JTX-666; Tr. 717:22-718:18 (Page); Trial Tr. 1718:19-1720:25 (Knowles).)
George Muller et al., Thalidomide Analogs and PDE4 Inhibition (“Muller 1998”), is a 1998 article authored by one of the inventors of the '536 Patent, George Muller. (Trial Tr. 713:10-24 (Page). See generally JTX-69.) Muller 1998 describes Celgene's efforts to prepare several analogs of thalidomide and increase their ability to inhibit TNFa and PDE4 in vitro as well as decrease their teratogenic potency. (JTX 69 1-2; Trial Tr. 713:25-714:20.) Muller reviews several thalidomide analogs and concludes that “these thalidomide analogs are potent inhibitors of PDE4” and “control TNFa levels by inhibition of PDE4.” (JTX 69 2, 5.)
The Court finds that the foregoing prior-art references, combined with the '358 Patent, do not make claim 6 of the '536 Patent obvious to the POSA “at the time the invention was made.” Mitsubishi, 2021 WL 1845499, at *11 (emphasis omitted) (quoting pre-AIA version of 25 U.S.C. § 103). Once again, to invalidate the '536 Patent for obviousness, Defendants must demonstrate that “a person having ordinary skill in the art would have had ‘reason to attempt to make the composition'” known as stereomerically pure apremilast to treat psoriasis and “a reasonable expectation of success in doing so.” Procter & Gamble Co. v. Teva Pharm. USA, Inc., 566 F.3d 989, 995 (Fed. Cir. 2009) (citation omitted). For the reasons discussed above regarding the '638 Patent, the Court finds that Defendants have failed to demonstrate that with the limited public data available to a POSA in 2002 on the feasibility of treating psoriasis with any of the '358 Patent example compounds, the POSA would have had a reason or motivation to attempt to separate or synthesize the compounds' enantiomers, let alone a reason to try and separate or synthesize the enantiomers of Example 12 to invent such a treatment. (But see Trial Tr. 1487:17-1488:2 (Davies) (“[I]f your racemate has some bad effects that you can't get around, then maybe it's a good idea to resolve it, if you know some information about the racemate that is bad for you. But if the racemate's fine, you would just carry on with the racemate.”).) Even if a POSA would have been motivated to try to isolate and study apremilast to treat psoriasis based on the prior art discussed above, the POSA would not have had a reasonable expectation of success in doing so without undue experimentation. Similarly, the Court finds that the objective indicia of non-obviousness weigh in Amgen's favor for the reasons discussed above regarding the '638 Patent.
C. The '541 Patent
The Court reaches a different result, however, for the '541 Patent. The '541 Patent issued on October 9, 2018 from U.S. Patent Application No. 14/826, 027, filed on August 13, 2015 and has an effective filing date of August 15, 2014. (Stipulation of Facts ¶¶ 63-66.) To show obviousness, Defendants must prove by clear and convincing evidence that “the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains.” 35 U.S.C. § 103(a) (post-AIA).
Amgen asserts claims 2, 19, and 21 of the '541 Patent against Defendants. (Stipulation of Facts ¶ 67.) Claim 2 of the '541 Patent recites as follows:
Before trial, the parties stipulated that if the '541 Patent was valid, Defendants' ANDA products would infringe claims 2, 19, and 21 of the '541 Patent. (Stipulation of Facts ¶¶ 14-15, Exhibit A2 of the Final Pretrial Order, ECF No. 422-1; Stipulation of Facts ¶¶ 9-12, Exhibit A3 of the Final Pretrial Order, ECF No. 422-2.)
[a] method of treating a patient with stereomerically pure (+)-2-[l-(3-ethoxy-4-methoxyphenyl)-2-methylsulfonylethyl]-4-acetylaminoisoindoline-1, 3-dione, wherein the patient is suffering from psoriasis, the method consisting of: (a) administering to a patient stereomerically pure (+)-2-[l-(3-ethoxy-4-methoxyphenyl)-
2-methylsulfonylethyl]-4-acetylaminoisoindoline-l, 3-dione in an initial titration dosing schedule consisting of (i) 10 mg in the morning on the first day of administration; (ii) 10 mg in the morning and 10 mg after noon on the second day of administration; (iii) 10 mg in the morning and 20 mg after noon on the third day of administration; (iv) 20 mg in the morning and 20 mg after noon on the fourth day of administration; (v) 20 mg in the morning and 30 mg after noon on the fifth day of administration; and (b) on the sixth and every subsequent day, administering to the patient 30 mg in the morning and 30 mg after noon of stereomerically pure (+)-2-[l-(3-ethoxy-4-methoxyphenyl)-2-methylsulfonylethyl]-4-acetylaminoisoindoline-l, 3-dione.(Id. ¶ 70.) Defendants assert that these claims are directed to a method of treating a patient's psoriasis according to the following twice daily apremilast dosing schedule (which culminates in a 30 mg, twice-per-day maintenance dose):
(Defs.' Post-Trial Br. 3.)
Day 1
Day 2
Day 3
Day 4
Day 5
Day 6
a.m.
10 mg
10 mg
10 mg
20 mg
20 mg
30 mg
p.m.
10 mg
20 mg
20 mg
30 mg
30 mg
Claim 19 of the '541 Patent recites as follows: “[a] method as in any one of claims 1-14, wherein the stereomerically pure (+)-2-[l-(3-ethoxy-4-methoxyphenyl)-2-methylsulfonylethyl]-4-acetylaminoisoindoline-1, 3-dione comprises greater than about 98% by weight of the (+) isomer of 2-[l-(3-ethoxy-4-methoxyphenyl)-2-methylsulfonylethyl]-4-acetylaminoisoindoline-l, 3-dione based on the total weight of 2-[l-(3-ethoxy-4-methoxyphenyl)-2-methylsulfonylethyl]-4-acetylaminoisoindoline- 1, 3-dione.” (Stipulation of Facts ¶ 71.) Claim 21 of the '541 Patent recites as follows: “[a] method as in any one of claims 1-14, wherein the stereomerically pure (+)-2-[l-(3-ethoxy-4-methoxyphenyl)-2-methylsulfonylethyl]-4-acetylaminoisoindoline-1, 3-dione is administered in tablet form.” (Id. ¶ 72.) There is little dispute between the parties as to the obviousness of these claims. Considering preceding patents such as the '536 Patent, which taught a method of treating psoriasis using stereomerically pure apremilast in tablet or capsule form as either a single or divided dose (JTX-720-21; Trial Tr. 833:22-834:2), the Court finds these claims obvious based on the analysis of claim 2 in the '541 Patent that follows.
Defendants challenge the validity of the '541 Patent, arguing that prior art “disclosed a fixed dosing titration schedule for apremilast up to a maintenance dose of 60 mg per day within five days.” (Defs.' Post-Trial Br. 6.) “In general terms, dose titration refers to the process of initiating treatment at a given dose” but adjusting the dose over a period of time “until a desired target dose is reached.” (Trial Tr. 1753:20-1754:3 (Alexis).) “Typically, ” physicians initiate treatment at a lower dose before “gradually increasing that dose, ” although the concept of dose titration is “agnostic to direction.” (Trial Tr. 1753:23-1754:1 (Alexis).) Defendants assert that the prior art teaches that “dose titration can be used to mitigate adverse events.” (Defs.' Post-Trial Br. 7 (citations omitted).) Defendants maintain that a POSA would have been motivated to modify prior art to arrive at the six-day titration schedule disclosed by the '541 Patent and to do so with a reasonable expectation of arriving at a more effective psoriasis treatment. (Id. at 19-21.) Defendants, accordingly, assert that the '541 Patent claims were obvious. (Id. at 21.) The Court agrees.
1. Graham Framework: Level of Ordinary Skill in the Art
The Court adopts Amgen's proposed definition of a POSA for purpose of the '541 Titration Patent. (See Pl.'s Proposed Findings of Fact ¶ 1545 (“The person of ordinary skill in the art for purposes of the '541 Titration Patent would have been a physician with training and experience in treating psoriasis or psoriatic arthritis, including experience evaluating clinical study results, and with access to someone with expertise in chemistry.” (citing Trial Tr. 1751:19-1752:3 (Alexis))).) Defendants maintain that their proposed POSA definition is not materially different from Amgen's. (Defs.' Proposed Findings of Fact ¶¶ 1800-01 (citing Trial Tr. 831:5-14 (Gilmore)).)
2. Graham Framework: Scope and Content of the Prior Art
a. Papp 2012
A 2012 Lancet article entitled “Efficacy of Apremilast in the Treatment of Moderate to Severe Psoriasis” (“Papp 2012”) is relevant, close prior art to the '541 Patent. (DTX-1531; Trial Tr. 832:1-10, 834:22-835:7 (Gilmore).) The article reports the results of a clinical trial investigating the “clinical efficacy and safety of different doses of apremilast in the treatment of patients with moderate to severe plaque psoriasis.” (DTX-153 1.) Evidence at trial established that the named inventor of the '541 Patent, Robert Day, is listed as one of the authors of Papp 2012. (Compare DTX-153J, with JTX-132, and Trial Tr. 1825:13-25.) The study reported by Papp 2012 was also funded by Celgene. (DTX-1531.) After investigating 10 mg, 20 mg, and 30 mg twice daily target doses of apremilast, Papp 2012 reports that apremilast, “given orally at 20 mg or 30 mg twice daily, seems to be efficacious, safe, and tolerable for patients with moderate to severe plaque psoriasis.” (DTX-1531.) In terms of the efficacy of a 20 mg dose versus a 30 mg dose, Papp 2012 concludes that “[a]lthough no statistical comparisons between apremilast doses were done, apremilast 30 mg had the most favourable outcome and this dose is being investigated for patients with moderate to severe plaque psoriasis.” (DTX-1537.) Based on the foregoing, the Court finds that Papp 2012 teaches a POSA that a 30 mg twice-daily maintenance dose produces the most favorable outcome for patients with moderate to severe plaque psoriasis.
In initiating apremilast treatment, Papp 2012 reports that patients' doses were “titrated in the first week to mitigate potential dose-dependent adverse events of apremilast; all patients reached the target dose by day 5.” (DTX-1532.) As discussed above regarding Otezla, see, supra, n.3, apremilast, Otezla's active ingredient, is associated with significant gastrointestinal-related side-effects such as nausea and diarrhea. (DTX-153J; DTX-1621.) And as the parties' experts explained at trial, “potential dose-dependent adverse events” means “as the dose increases, sometimes patients have more side effects.” (Trial Tr. 839:6-12 (Gilmore), 1828:19-1829:4 (Alexis) (testifying that a dose-dependent adverse event means “there's a dose response relationship in the frequency or severity of adverse events; meaning that if the dose increases, the likelihood of an event would increase too”).) It is generally known in the art that a POSA can “try to mitigate those side effects by slowly increasing the dose of the medication.” (Trial Tr. 839:6-14 (Gilmore).) The following chart summarizes the twice-daily 30 mg target dose titration schedule disclosed by Papp 2012:
(See Trial Tr. 840:14-841:3 (Gilmore), 1835:21-25 (Alexis) (admitting that “this is the dosing schedule used in the clinical trial that Papp 2012 published in a manuscript in 2012 about in the Lancet”), 1836:9-18 (agreeing that on days 1-2 of the clinical trial protocol underlying Papp 2012, patients were administered 10 mg in the morning and 10 mg in the evening, for a total dose of 20 mg per day, on days 3-4 patients were administered 20 mg in the morning and 20 mg in the evening, for a total dose of 40 mg per day, and on day 5 patients were administered the maintenance dose of 30 mg twice a day, for a total dose of 60 mg per day).) Notwithstanding this dose titration schedule, Papp 2012 reports that 18% of patients in the 30 mg twice-daily treatment group experienced nausea, 16% experienced upper respiratory tract infections, 14% experienced diarrhea, and 16% experienced tension headaches. (DTX-1536 (Table 2); Trial Tr. 848:7-25 (Gilmore).)
1st Dose (mg)
2na Dose (mg)
Total dose (per day)
Day 1
10
10
20
Day 2
10
10
20
Day 3
20
20
40
Day 4
20
20
40
Day 5
30
30
60
Day 6 (and thereafter)
30
30
60
b. Schett 2012
A 2012 article entitled “Oral Apremilast in the Treatment of Active Psoriatic Arthritis: Results of a Multicenter, Randomized, Double-Blind, Placebo-Controlled Study” (“Schett 2012”), is also relevant close prior art to the Court's obviousness analysis. (See generally DTX-162.) Schett 2012 reports on a study of apremilast where patients were given a target dose of “40 milligrams per day, which could have been given as a single dose, or as 20 milligrams twice per day.” (DTX-1621; Trial Tr. 1776:1-5 (Alexis), 850:5-11 (Gilmore).) Although not a study of plaque psoriasis patients, like Papp 2012, the study described by Schett 2012 implemented a “[d]ose escalation” protocol “during the first 7 days of treatment in an attempt to decrease the likelihood of adverse events (AEs) related to treatment initiation.” (DTX-1622; Trial Tr. 850:16-23.) The protocol associated with Schett 2012 provides additional detail on the study's dose escalation, disclosing that in the first week of the study, patients received a fixed dose titration of 10 mg once a day for days one through three followed by 20 mg once a day for days four through seven, during the first week of dosing to “ameliorate the dose dependent adverse events of [apremilast] (headache and GI disturbances).” (DTX 923; Trial Tr. 850:19-854:9, 857:7-17.)
c. Pathan 2012
Finally, the Court finds that the 2012 article entitled “Efficacy and Safety of Apremilast, an Oral Phosphodiesterase 4 Inhibitor, in Ankylosing Spondylitis” is relevant prior art to the '541 Patent. (See generally DTX-157.) In that study, patients “were started on apremilast 10 mg twice daily or placebo and the dose was titrated by 20 mg every 2 days until the maximum dose of 30 mg twice daily was achieved on day 5.” (DTX-1572; Trial Tr. 845:20-846:4 (Gilmore), 1840:7-18 (Alexis).)
3. Graham Framework: Differences between Claims and Prior Art
Here, the Court is confronted with close prior art, each teaching fixed dosing schedules with some version of a twice-daily dosing option. Papp 2012 taught a POSA that a twice-daily 30 mg apremilast maintenance dose produces the most favorable outcome for patients with moderate to severe plaque psoriasis. Papp 2012 also taught a five-day schedule for reaching that dose. Schett 2012 taught that initiating apremilast treatment with a 10 mg dose of the drug before titrating up in 10 mg increments during the first week of treatment is a strategy for ameliorating dose dependent adverse events. The evidence at trial also established that none of the peer reviewed literature teaches a dosing titration schedule that extended beyond eight days before reaching the target apremilast dose. (Trial Tr. 1845:15-23 (Alexis).) Moreover, evidence at trial established that it was well within a POSA's ability to titrate an apremilast dose for a patient presenting with psoriasis, and that doing so is a “routine aspect of treating psoriasis with any of those drugs that require dose titration.” (Trial Tr. 858:11-20 (Gilmore), 1812:15-1813:22 (Alexis).) On these facts, the Court finds that a twice-daily, six-day, fixed apremilast dosing schedule, initiating treatment at 10 mg and increasing toward a 30 mg target dose in 10 mg increments is rendered obvious by the prior art. Cf. KRS Int'l Co., 550 U.S. at 417 (“If a person of ordinary skill can implement a predictable variation, § 103 likely bars its patentability.”).
As Defendants argue, Hoffman-La Roche is strikingly similar to this case. In that case, the Federal Circuit considered “whether it would have been obvious at the time of invention to select a once monthly oral dosing regimen” of a drug to treat a disease, and “whether it would have been obvious to set that dose at 150 mg.” Hoffman La Roche v. Apotex, 748 F.3d 1326, 1329 (Fed. Cir. 2014). To treat the disease with the drug, a tablet needed “to be taken in a fasting state at least 30 minutes before eating or drinking.” Id. at 1328. This requirement was an “inconvenience” that “created problems of patient compliance.” Id. “Researchers in the field believed that less-frequent dosing would result in patients continuing the treatment for the long term, which is required for bisphosphonate treatments to be successful.” Id.
At the same time, the prior art suggested that a less frequent cumulative monthly dose rather than continuous daily dosing could successfully treat the disease. Id. at 1330-31. Furthermore, among the studied daily doses, “only the 2.5 and 5 mg doses showed positive outcome in all regions.” Id. at 1332. Although “the 5 mg dose did not demonstrate greater efficacy than the 2.5 mg dose, it was still deemed an equivalently effective dose so that someone scaling it to a single monthly dose of 150 mg (5 mg/day x 30 days/month) would have anticipated equivalent success” in treating the disease. Id. Accordingly, the Federal Circuit found that “the prior art pointed to a monthly treatment of 150 mg of ibandronate. At the very least, the 150 mg dose was obvious to try: There was a need to solve the problem of patient compliance by looking to less-frequent dosing regiments.” Id. The Court noted that although the efficacy of a once monthly dose at 150 mg had never been precisely studied in the prior art, “[c]onclusive proof of efficacy is not necessary to show obviousness. All that is required is a reasonable expectation of success.” Id. at 1331.
Similarly, based on the record before the Court, it appears that the efficacy of a six-day, fixed titration schedule for an apremilast treatment for moderate to severe plaque psoriasis treatment had not been studied in the prior art. Nevertheless, as Amgen's expert acknowledged, the POSA would have had a reasonable expectation of success treating psoriasis by titrating up the apremilast dose to 30 mg given twice daily on the sixth day of treatment. (Trial Tr. 859:4-12 (Gilmore), 1848:8-25 (Alexis) (admitting that the “efficacy of the maintenance dose of 60 milligrams a day, we are informed of that efficacy by the Papp 2012 article. If the titration schedule is extended by one day ... I personally would not have the expectation that it would not have efficacy”).) Also like Hoffman-La Roche, in selecting a target dose of 30 mg twice daily, adding an additional day to the Papp 2012 schedule and titrating up in 10 mg increments per day instead of Papp 2012' s 20 mg increments, were obvious solutions to try because “there were only a ‘finite number of identified, predictable solutions'” in the prior art for managing tolerability concerns while also delivering efficacious treatment. 748 F.3d at 1332 (quoting KRS Int'l Co., 550 U.S. at 421); see also In re Copaxone Consol. Cases, 906 F.3d 1013, 1025 (Fed. Cir. 2018) (“Here, the prior art focused on two critical variables, dose size and injection frequency, and provided clear direction as to choices likely to be successful in reducing adverse side effects and increasing patient adherence.”). As Dr. Gilmore explained, titrating up by 10 mg per day instead of 20 mg per day as used in Papp 2012 “is like tak[ing] the stairs one at a time instead of two at a time” (Trial Tr. 861:19-863:7 (Gilmore)), an obvious way to avoid the strain of dose dependent side effects while taking a longer time to arrive at the ultimate target dose.
And plainly, in prescribing a drug with dose-dependent adverse events in the early weeks of treatment, the POSA would have been motivated to try to extend the Papp 2012 schedule and titrate up in smaller amounts. Dr. Gilmore credibly testified that many of her patients “decided not to continue with treatment” because of gastrointestinal-related side effects from apremilast. (Trial Tr. 877:11-23 (Gilmore).) Relatedly, evidence at trial suggested that Celgene was concerned that “GI tolerability is a major clinical barrier to starting and keeping patients on Otezla.” (JTX-100 25.) Because the prior art provides dose titration as a method of reducing these side effects, a POSA would have been motivated to look to a six-day titration schedule as a strategy for increasing patient compliance.
Amgen has elected not to present objective indicia of non-obviousness. (Pl.'s Post-Trial Br. 89 n.33 (“Amgen cannot be faulted for choosing not to put forward evidence of objective indicia supporting non-obviousness of the '541 Patent claims in order to focus its trial presentation.”).)
Based on the foregoing, the Court finds that claims 2, 19, and 21 of the '541 Patent are invalid as obvious.
D. The '101 Form B Patent
Amgen also asserts that DRL, Sandoz, and Zydus's proposed ANDA products infringe claims 1 and 15 of the '101 Patent. Claim 1 recites “[a] form B crystal form of the compound of Formula (1):
(Image Omitted) which is enantiomerically pure, and which has an x-ray powder diffraction pattern comprising peaks at about 10.1, 13.5, 20.7, and 26.9 degrees 20.” (Stipulation of Facts ¶ 50.) Claim 15 recites “[a] solid pharmaceutical composition comprising the crystal form of any one of the claims 1 and 2 to 13.” (Id. ¶¶ 47, 51.)
1. Infringement of the '101 Patent
In response to Amgen's infringement assertions, Zydus maintains that Amgen “has failed to establish that, more likely than not, Zydus's proposed ANDA product, as it would be made, used, offered for sale, or sold in the United States or imported into the United States, will contain crystalline apremilast Form B.” (Defs.5 Post-Trial Br. 77; see Glaxo, Inc. v. Novopharm, Ltd., 110 F.3d 1562, 1570 (Fed. Cir. 1997) (“The relevant inquiry is whether the patentee has proven by a preponderance of the evidence that the alleged infringer will likely market an infringing product.”).) Here, the Court agrees with Zydus that Amgen has failed to meet its burden of demonstrating infringement.
Both DRL and Sandoz stipulate that submission of their ANDAs to the FDA seeking approval and subsequent manufacture and sale of their ANDA products will infringe claims 1 and 15 of the '101 Patent if those claims are not found to be invalid or unenforceable. (DRL and Sandoz Stipulation of Facts ¶¶ 9-12.)
The parties do not dispute that “[a]premilast has several crystal forms, or polymorphs, including a polymorph referred to as ‘Form B.'” (Zydus Stipulation of Facts ¶ 16.) Polymorphism occurs when a given atom or molecule can crystallize into more than one crystalline form. (Trial Tr. 426:5-427:3 (Myerson).) The specific crystalline form of an active pharmaceutical compound may impact its pharmaceutical properties, including bioavailability, manufacturability, and shelf-life. (Trial Tr. 428:9-23 (Myerson), 1075:5-21 (Steed), 1169:19-1170:3 (Sacchetti).)
Synchrotron X-ray Powder Diffraction (“XRPD”) is an analytical technique for distinguishing polymorphic forms. (Trial Tr. 1221:11-1222:7 (Miller).) In this process, x-rays are shone on a powder sample of interest and the x-rays diffract based on how the powder's molecules are arranged into crystals. (Trial Tr. 428:25-429:20 (Myerson).) Diffraction is measured with a diffractometer on a detector. (Trial Tr. 429:11-16 (Myerson).) A “diffractogram” is a graphical depiction of the measured x-ray diffraction when shone on the powder sample. (Trial Tr. 429:7-20 (Myerson).) “Each one of the peaks” in a diffractogram's “diffraction pattern is a direct function of the lattice spacing and orientation of the lattice planes in the crystal structure. And there are many lattice planes in a crystal structure.” (Trial Tr. 1222:20-23 (Miller).) As Zydus's polymorph expert, Dr. Steven Miller, explained, “[e]ach polymorph has its own unique x-ray powder diffraction pattern.” (Trial Tr. 1138:18-19 (Miller).) And because some of the diffractogram's peak locations are unique, “diffraction patterns with relation to polymorphic forms” can operate similar to a “fingerprint” analysis. (Trial Tr. 1223:7-17 (Miller).) Dr. Miller credibly explained that in performing this analysis it is important to try and “identify as many of the peaks or match as many of the peaks as possible, ” so as to have multiple points of comparison. (Trial Tr. 1223:7-24 (Miller).) As Dr. Miller testified, like fingerprints, “we certainly wouldn't want to, you know, convict somebody of a single point of match ... in a fingerprint.” (Trial Tr. 1223:23-24 (Miller).)
The parties also do not contest that Zydus's proposed ANDA product and its apremilast active pharmaceutical ingredient (“API”) includes Form A apremilast. (Trial Tr. 439:9-21 (Myerson).) This version of apremilast differs from the Form B version of apremilast at issue in claims 1 and 15. (Trial Tr. 1248:3-1249:3 (Miller).) Although Amgen asserts that Zydus's ANDA product also includes Form B apremilast generating certain claimed peaks, proving this assertion via a diffractogram is complicated because “it is possible to get mixtures of polymorphic forms, or of any different, you know, crystalline compound. And then, you have to think of that like as overlapping fingerprints, which obviously complicates the matter quite a bit.” (Trial Tr. 1224:1-5 (Miller).)
Based on its experts' analysis, Amgen introduced the following diffractogram that it asserts demonstrates that Zydus's API generates Form B XRPD peaks at 10.1 and 13.5 ± 0.2° 29, thereby allegedly infringing the '101 Patent. In the following diffractograms, the green patterns represent Zydus's proposed ANDA product, the black lines represent Zydus's API, the red lines represent the Form A reference standard, and the blue patterns represent the Form B reference standard.
Color versions of the diffractograms appear in the Court's slip opinion.
(Image Omitted) (Trial Tr. 1242:4-1243:15 (Miller); PTX-1238 38 (Gozzo Diffractograms).)
As Dr. Miller convincingly explained, however, both Form A and B peaks at 10.1 and 13.5 overlap with each other. (Trial Tr. 1241:9-1242:2 (“[W]e actually do have peaks roughly at that position of 10.1 degrees. However, you can see that these peaks coincide with the red diffractogram for the Form A reference standard. No one at this point has disputed that Form A is present in these samples.”), 1242:22-1243:19 (Miller) (providing a similar explanation as to the peak at 13.5 degrees, noting that in the blue line representing the Form B reference standard, there is a “doublet, ” featuring two peaks right next to each other near 13.5 degrees, which is not present in the samples, thus “support[ing] the notion that the Form A is present and that's what's generating that peak at 13.5 degrees.”).) Accordingly, without more, a diffractogram demonstrating the presence of both of these peaks does not prove by a preponderance of evidence the presence of Form B as opposed to Form A. This is especially true where, as here, it is undisputed that Form A is present in Zydus's ANDA product and API.
Amgen also produced the following evidence as to the claimed 20.7 and 26.9 ± 0.2° 20 peaks:
(Image Omitted) (Miller Tr. 1233:17-1236:12; PTX-123839 (Gozzo Diffractograms).) The Court finds persuasive Dr. Miller's testimony that “for 20.7 and 26.9, there are, in fact, no peaks actually present.” (Trial Tr. 1232.24-1233.10 (Miller).) As for the peak at 26.9, on cross-examination, Amgen's expert Dr. Myerson appeared to agree with Zydus's counsel that “the peak list from the Zydus sample shows no peak at 26.96.” (Trial Tr. 481:16-23 (Myerson) (“Right.”).) Although Dr. Myerson attempted to explain the absence of that peak by opining “that peak is probably not resolvable because of the broadness of the big the big peak at 26.880, ” Dr. Myerson did little to explain the import of the alleged “broadness” of a neighboring peak. (Trial Tr. 480:18-23 (Myerson).)
As for the peak at 20.7, Dr. Myerson acknowledged that that peak was absent from the Zydus samples. (Trial Tr. 483:1-3 (“Q: And if you look at those peak lists, there [are] no peaks within plus or minus .2 of 20.7, correct? A: That's correct.”).) Dr. Myerson attempted to explain away this peak's absence by referencing a concept in XRPD testing known as “limit of detection.” (Trial Tr. 487:18-488:6 (Myerson).) In Dr. Myerson's explanation, the 20.7 peak could not be measured by the diffractometer “because of background noise in a diffractogram.” (Trial Tr. 487:18-488:6 (Myerson).) Dr. Miller offered his own definition of limit of detection as “the smallest detectable amount of a particular substance given the instrumentation and procedures used” and “the number at which you can distinguish its detectability and the absence of a substance.” (Trial Tr. 1238:10-23 (Miller).)
But as Dr. Miller explained, “a peak is either there or it is not there. . . . [T]he test either results in a peak, or it doesn't result in a peak.” (Trial Tr. 1241:4-6 (Miller).) And as Dr. Miller further explained, “if nothing shows up in the test, i.e., a peak in the case of XRPD, then the assumption is that it's not there.” (Trial Tr. 1239:11-13 (Miller).) Moreover, as Zydus notes, Dr. Myerson's assertion that the 20.7 peak is below the limit of detection is especially bald considering his failure to opine on what the limit of detection may have been for the 20.7 Form B peak in the context of Dr. Gozzo's experiments. (Trial Tr. 480:18-23 (Miller).) And in all events, it is Amgen's burden to demonstrate the existence of the claimed '101 Patent peaks in Zydus's samples. Glaxo, 110 F.3d at 1565-66 (affirming a district court's finding that a plaintiff failed to meet its infringement burden where it failed to show that the defendant's product “met substantially all of the limitations of' a “claimed 29-peak spectrum”).
Finally, Zydus presented evidence at trial of its own internal polymorph testing of its apremilast API and proposed ANDA product. (Trial Tr. 1256:7-25 (Miller); JTX-9003-18; JTX-90115-21; JTX-9023-31.) Dr. Miller credibly testified that based on his review of these reports, Zydus's internal testing shows that its sample API and ANDA product contain Form A and do not contain Form B apremilast. (Trial Tr. 1256:7-1257:10 (Miller).) Based on his extensive experience with XRPD testing in the pharmaceutical context, including his experience analyzing approximately 50 XRPD patterns per week, the Court finds Dr. Miller's testimony persuasive in this regard. (Trial Tr. 1216-20, 1219:1-1220:11 (Miller).)
The Court, accordingly, finds that Amgen has failed to meet its burden of proving Zydus's infringement of the '101 Patent.
Because the Court finds that Amgen has failed to prove Zydus's infringement, the Court declines to consider Zydus's alternative argument that Amgen's infringement case fails because it allegedly tested unrepresentative ANDA and API samples. Accordingly, Zydus's Motion for a Directed Verdict as to claims 1 and 15 of the '101 Patent claims (Trial Tr. 515:7-9) is denied as moot.
2. Obviousness and the '101 Patent
The Court next considers whether the '101 Patent claims are invalid for obviousness. To start, the parties dispute the priority date to which the '101 Patent is entitled. Amgen maintains that the claims asserted in the '101 Patent are entitled to a March 2002 priority date. (Pl.'s Post-Trial Br. 68.) According to Amgen, the '101 Patent issued from a continuation in part application that cited a March 20, 2002 '515 Provisional Patent on its face. (Id.; see also JTX-52.) The '515 Provisional Patent, Amgen argues, “provides written description and enablement support for claims 1 and 15.” (Pl.'s Post-Trial Br. 68-69.) Amgen invokes the doctrine of inherent disclosure in arguing that the '515 Provisional Patent satisfies the written description requirement for claims 1 and 15. (Id. at 69.) For their part, Defendants assert that Amgen is not entitled to the March 20, 2002 priority date because it cannot satisfy the prerequisites for inherent disclosure. (Defs.' Post-Trial Br. 71.) Accordingly, Defendants argue, the obviousness of Amgen's '101 Patent claims must be considered as of March 27, 2008, the undisputed date on which the '101 Patent was filed. (Id. at 63.) For the reasons set forth below, the Court agrees that Amgen is entitled to the March 2002 priority date.
The Federal Circuit has held that under the “doctrine of inherent disclosure, when a specification describes an invention that has certain undisclosed yet inherent properties, that specification serves as adequate written description to support a subsequent patent application that explicitly recites the invention's inherent properties.” Yeda Rsch. & Dev. Co. v. Abbott GMBH & Co. KG, 837 F.3d 1341, 1345 (Fed. Cir. 2016). “‘Inherent' properties, . . . are the rare exceptions to the rule that a party must show possession of ‘every feature' recited in the count and that ‘every limitation' of the count must have been known to the inventor at the time of the alleged conception.” Hitzeman v. Rutter, 243 F.3d 1345, 1354-55 (Fed. Cir. 2001) (citation omitted). Inherency “requires that the missing descriptive material is necessarily present, not merely probably or possibly present, in the prior art.” Trintec Indus., Inc. v. Top-U.S.A. Corp., 295 F.3d 1292, 1295 (Fed. Cir. 2002) (internal quotation marks omitted). “[I]f the teachings of the prior art can be practiced in a way that yields a product lacking the allegedly inherent property, the prior art in question does not inherently anticipate.” In re. Armodafinil Patent Litig. Inc., 939 F.Supp.2d 456, 465 (D. Del. 2013) (citing Glaxo Inc. v. Novopharm Ltd., 52 F.3d 1043, 1047-48 (Fed. Cir. 1995) (finding no inherency where prior art example could yield crystals of either the claimed polymorph or a different polymorph)).
Here, the parties do not dispute that the '515 Provisional Patent filed on March 20, 2002 discloses a synthetic chemical procedure for preparing apremilast identified as Example 2. (See Trial Tr. 1576:15-1577:1 (with Dr. Myerson testifying that Example 2 is a “synthetic procedure, a chemical procedure .... It's a recipe for preparing Compound A”), 1572:25-73:1 (“Example 2 is entitled ‘Preparation of Compound A,' Compound A being apremilast”), 1088:2-7 (Steed) (“[Example 2] is the recipe, if you like, by which the apremilast molecule can be made”); JTX-4329.) Example 2 does not, however, explicitly describe the final resulting form of apremilast as a crystalline structure-only as a solid. (Trial Tr. 1093:12-1094:1, 1110:10-12 (Steed) (“Example 2 of that '515 application does not explicitly disclose a crystalline solid of apremilast. All we know is that it's a solid”); Trial Tr. 1576:4-7 (Myerson) (“[A]ll experts agree that a POSA would have understood on March 20th, 2002, that Example 2 results in some solid form of apremilast.” (emphasis added)).) Nevertheless, as Dr. Steed explained at trial, in chemistry, “typically there are two kinds of solid forms .... They can either be crystalline or amorphous.” (Trial Tr. 1072:25-1073:4 (Steed).) Indeed, as Defendants note, the '101 Patent itself provides that apremilast may be found in seven different crystalline forms and one amorphous form. (See JTX-563, at 52:57:60; Trial Tr. 1088:25-1089:12.)
In performing its obviousness analysis of the' 101 Patent, the Court accepts Amgen's definition of the POSA as an individual with a bachelor's degree in chemistry, chemical engineering, or a related discipline, with some knowledge of crystalline solid forms and their characterization, and several years of experience in the pharmaceutical industry; or an advanced degree in the above listed disciplines, with some knowledge of solid-state chemistry or analytical chemistry and less experience. (Pl.'s Proposed Findings of Fact ¶ 1204.) Defendants assert that their proposed definition of a POSA for purposes of the '101 Patent is not “materially different” from Amgen's. (Defs.' Proposed Findings of Fact ¶¶ 1400-02.)
The final step of Example 2 provides that “the residue recrystallized from a binary solvent containing ethanol (150 mL) and acetone (75 mL).” (JTX-4329; Trial Tr. 1596:1-7 (Myerson).) Amgen asserts that from this disclosure, a POSA would have understood that the '515 Provisional Patent inventors were in possession of a crystalline form of apremilast. (Pl.'s Post-Trial Br. 70.) At trial, Amgen's expert, Dr. Myerson, testified that “Example 2 inherently produces crystalline Form B whenever followed by a POSA.” (Trial Tr. 1576:8-9 (Myerson).) As Defendants note, however, Dr. Myerson did not cite any disclosure in Example 2 or within the '515 Provisional Patent application to support the assertion that the '515 Provisional Patent inventors would have produced Form B apremilast from Example 2. (Trial Tr. 1623:21-1624:22, 1625:14-1626:2 (Myerson) (noting that he did not “have any data that would demonstrate what form [Celgene] made, ” he hasn't “seen any lab notebooks” to determine what form Celgene made, and has seen no analytical data for the product of Example 2 in the '515 application). On the other hand, Dr. Myerson cited thirteen experiments performed by Teva, Zentiva, and Lek that replicated the procedures outlined in Example 2. (Trial Tr. 1577:25-1578:17; JTX-225l-2.) The results of these replications were ultimately submitted to the European Patent Office (“EPO”) as part of a proceeding unrelated to this case. (Trial Tr. 1577:25-1578:17 (Myerson).) According to Dr. Myerson, each experiment performed by these companies replicated Example 2 and resulted in Form B. (Trial Tr. 1587:18-19 (Myerson).)
In response to the Example 2 replications submitted to the EPO, Defendants assert that “evidence presented at trial showed that following Example 2 does not always lead to the claimed apremilast crystalline Form B-rather it could lead to Form A or Form C.” (Defs.' Post-Trial Br. 72.) Indeed, Defendants introduced evidence at trial that Celgene itself represented to the EPO that following Example 2 would lead to apremilast crystalline Form C, not Form B. (JTX-22514-15; Trial Tr. 1106:6-16; 1108:2-9 (Steed) (“Celgene itself represented to the [EPO] that following the teachings of Example 2 including that recrystallization step results in crystalline Form C . . . not Form B. And Celgene is the patent [proprietor], so they should know what they're doing.”), 1108:17-1109:8 (explaining that Celgene submitted data from experiments that Celgene maintained replicated the Example 2 recrystallization step in three different ways, all of which led to Form C); see also Trial Tr. 1620:1-6 (Myerson) (acknowledging the same).) Notwithstanding this representation to the EPO, however, Dr. Myerson's uncontradicted trial testimony was that although Celgene once claimed that it made Form C following Example 2, Form C involves the use of toluene solvate, which becomes a part of the crystal structure that makes up Form C. (Trial Tr. 1586:18-1587:4 (Myerson).) Nevertheless, toluene is not mentioned in Example 2 of the '515 Provisional Patent. (Trial Tr. 1587:6-10 (Myerson).) The Court, accordingly, finds persuasive Dr. Myerson's testimony that “Celgene just made a mistake” in representing that Example 2 could produce Form C apremilast. (Trial Tr. 1587:5 (Myerson).)
Defendants' expert Dr. Steed also disputed whether or not Example 2 inherently produced the Form B apremilast claimed in the '101 Patent. Dr. Steed testified that “dependfing] upon how the person of skill chose to do the recrystallization step . . . Form A can result from ethanol and acetone if fast crystallization is used, Form B from slow, and any one of the other crystalline forms could also result.” (Trial Tr. 1105:18-21 (Steed).) Amgen's expert Dr. Myerson, however, testified that in several of the Example 2 replications submitted to the EPO, the companies used fast cooling but nevertheless produced Form B apremilast. (See, e.g., Trial Tr. 1581:6-1582:3 (Myerson).) In Dr. Myerson's opinion, “[a]s we saw in the EPO data, if you're using a 2:1 ratio of ethanol to acetone, you use fast cooling or slow cooling, you always get Form B.” (Trial Tr. 1596:19-22 (Myerson).)
In deciding the question of the '101 Patent's priority date, the Court is mindful of each party's evidentiary burdens. Here, “for a patent's claims to be entitled to an earlier priority date, the patentee must demonstrate that the claims meet the requirements of 35 U.S.C. § 120.” Nat. Alt. InfI, Inc. v. lancu, 904 F.3d 1375, 1380 (Fed. Cir. 2018). “Accordingly, claims in a patent or patent application are not entitled to priority under § 120 at least until the patent owner proves entitlement to the PTO, the Board, or a federal court.” Id.; see also Tech. Licensing Corp. v. Videotek, Inc., 545 F.3d 1316, 1327 (2008) (requiring patentees filing continuation-in-part applications to “show not only the existence of the earlier application, but why the written description in the earlier application supports the claims”).
The Court finds that based on the testimony of Dr. Myerson and his descriptions of the thirteen successful replications of Example 2 submitted to the EPO, Amgen has shown that Example 2 of the '515 Provisional Patent inherently produces Form B apremilast as disclosed in the '101 Patent. Moreover, as Dr. Myerson explained, “[a] POSA would understand that the powder diffraction peaks, the XRPD peaks, ” recited in claims 1 and 15 of the '101 Patent “are an inherent property of Form B.” (Trial Tr. 1587:13-24.) Accordingly, Amgen has met its burden of showing that the '101 Patent is entitled to the '515 Provisional Patent's March 20, 2002 priority date. Yeda Rsch. & Dev. Co., 837 F.3d at 1345 (holding that under the “doctrine of inherent disclosure, when a specification describes an invention that has certain undisclosed yet inherent properties, that specification serves as adequate written description to support a subsequent patent application that explicitly recites the invention's inherent properties.”); Allergan, Inc., 796 F.3d at 1309 (“A claim that recites a property that is necessarily inherent in a formulation that is adequately described is not invalid as lacking written description merely because the property itself is not explicitly described.”).
Having demonstrated its entitlement to the earlier priority date, “the burden of going forward again shifts to the proponent of the invalidity defense, [Defendants], to convince the [C]ourt that [Amgen] is not entitled to the benefit of the earlier filing date.” Tech. Licensing Corp., 545 F.3d at 1328-29 (“[B]ecause an issued patent is by statute presumed valid, a challenger has the burden of persuasion to show by clear and convincing evidence that the contrary is true”). “‘Convince' is the operative word, because if the court is not persuaded by clear and convincing evidence that [Defendants are] correct, [Defendants have] failed to carry [their] ultimate burden of persuasion, and [their] defense of invalidity . . . fails.” Id. at 1328.
Defendants assert that “[e]ven if following Example 2 under certain circumstances using undisclosed parameters may lead to Form B, it does not follow that Example 2 would always result in Form B.” (Defs,' Post-Trial Br. 72; see Peshlakai v. Ruiz, No. 13-752, 2013 WL 6503604, at *14 n.6 (D.N.M. Nov. 20, 2013) (observing that an inference that “all swans are white” from past observations that “all observed swans have been white” is “vulnerable to a later specific observation that undermines the inferred conclusion-for example, later observing a black swan.” (citations omitted)). But Amgen's evidence has shifted the burden of production to Defendants and they must produce evidence that Form B is not inherent in Example 2. Although Defendants assert that where Example 2 is practiced with fast cooling, Form B does not necessarily result (Trial Tr. 1105:13-22 (Steed)), the Court is not convinced of this argument considering Dr. Myerson's rebuttal testimony based on his review of the data submitted to the EPO reproducing the process outlined in Example 2. Based on this review, Dr. Myerson persuasively testified that “if you're using a 2:1 ratio of ethanol to acetone, you use fast cooling or slow cooling, you always get Form B.” (Trial Tr. 1596:19-22.) And as noted above, the Court finds persuasive Dr. Myerson's testimony that to the extent Celgene represented to the EPO that Example 2 could yield Form C, that representation was mistaken. Accordingly, for these reasons, and upon review of the complete trial record, the Court finds that Defendants have failed to persuade the Court that Amgen's claim to the earlier '515 Provisional Patent priority date is unwarranted. Tech. Licensing Corp., 545 F.3d at 1327 (“Failure to prove the matter as required by the applicable standard means that the party with the burden of persuasion loses on that point-thus, if the [fact-finder] is left uncertain, the party with the burden loses.”).
Furthermore, Defendants fail to argue that art prior to March 2002 renders the '101 Patent invalid for obviousness. (See Defs.' Post-Trial Br. 63-74.) Indeed, as Amgen argues, it appears that “Defendants' obviousness arguments. . . reduce to a question of whether claims 1 and 15 of the '101 Patent are entitled to claim priority to March 20, 2002 through the '515 Provisional.” (Pl.'s Post-Trial Br. 68.) The Court, accordingly, finds that the '101 Patent is valid.
E. The '283 Form A Patent
Finally, the Court considers Zydus's assertion that the '283 Patent is invalid for both anticipation and obviousness. The '283 Patent was issued on January 10, 2012 from an application filed on November 12, 2010. (Stipulation of Facts ¶¶ 53, 56.) The FDA's Orange Book lists the expiration of the '283 Patent as March 19, 2023. In the instant action, Amgen asserts claims 2 and 27 of the '283 Patent against Zydus. (Id. ¶ 57.) Claim 2 of the '283 Patent recites “[t]he crystal form of claim 1, wherein the crystal form is Form A, which has an X-ray powder diffraction pattern comprising peaks at about 8.1, 14.4, 17.4, 23.6, and25.1 degrees 20.” (Id. ¶ 60.) For its part, claim 1 of the '283 Patent recites “[a]n unsolvated crystal form of the compound of Formula (I):
(Image Omitted) which is enantiomerically pure, wherein the crystal form is Form A, which has an X-ray powder diffraction pattern comprising peaks at 8.1, 14.4, 17.4, 23.6 and 25.1 degrees 20, or Form F, which has an X-ray powder diffraction pattern comprising peaks at about 8.1, 15.6, 17.3, and 25.4 degrees 20.” (Id. ¶ 59.) Claim 27 of the '283 Patent is an independent claim and recites: “[a] solid pharmaceutical composition comprising the crystal form of claim 2.” (Id. ¶ 61.)
1. Obviousness as to the '283 Form A Patent
To show obviousness, Zydus must prove by clear and convincing evidence that “the differences between the subject matter sought to be patented and the prior art are such that the subject matter as a whole would have been obvious at the time the invention was made to a” POSA. 35 U.S.C. § 103(a) (pre-AIA). Here, the parties stipulate to a March 27, 2008 priority date for the '283 Patent. (Stipulation of Facts ¶ 107.) Amgen argues that Zydus failed to show that a POSA would have had a reasonable expectation of success in obtaining Form A as of March 27, 2008. (Pl.'s Post-Trial Br. 77.)
Zydus's '283 obviousness combinations principally rely on U.S. Patent Application Publication No. U.S. 2003/0187052 Al (“the '052 Publication”), which was published on October 2, 2003. (Stipulation of Facts ¶ 153.) The uncontradicted trial testimony established that the '052 Publication is “what the '515 application eventually was published as.” (Trial Tr. 1113:9-10 (Steed).) The '052 Publication teaches that after the chemical synthesis of apremilast, the crude apremilast residue was purified using a 2:1 mixture of ethanol and acetone as solvents. (DTX-17915, ¶ [0103]); Trial Tr. 1112:12-1123:7 (Steed).) Furthermore, the '052 Publication discloses examples of preclinical testing performed on enantiomerically pure apremilast (DTX-17915-19 (Examples 3-8); Trial Tr. 1628:22-1629:18 (Myerson).) The '052 Publication teaches that incorporating enantiomerically pure apremilast into pharmaceutical compositions and single unit dosage forms is “useful in “treating and/or preventing disorders ameliorated by the reduction of levels of TNF-alpha or the inhibition of PDE4.” (DTX-1791 (abstract).)
Before the '283 Patent, recognizing the utility of stereomerically pure apremilast as disclosed in the '052 Publication, a POSA would have been motivated to perform a polymorph screen on the stereomerically pure apremilast disclosed in the '052 Publication. (Trial Tr. 1629:2-1630:13 (Myerson).) By that time, polymorph screening was a “routine task of pharmaceutical pre-formulation.” (Trial Tr. 1115:1-8 (Steed).) A polymorph screen typically includes a series of “experiments in which . . . the compound is crystallized under a variety of different conditions to see what solid forms ... it has.” (Trial Tr. 1077:1-9 (Steed).) These experiments can be as simple “as just dissolving the compound solution. It could even be water and letting it cool and crystalize and then analyzing the outcome.” (Trial Tr. 1077:1-9 (Steed).) The purpose of that screening would have been to “see what crystalline or amorphous solids [apremilast] actually forms. So, they could select amongst them which would be the most appropriate to formulate into a single unit dosage form.” (Trial Tr. 1114:17-1115:10 (Steed).) At trial, Amgen's expert, Dr. Myerson, agreed, testifying that a POSA “would be interested in finding a crystalline solid form that met a series of properties, one of which would be polymorph stability among other properties that are very important in choosing a solid form.” (Trial Tr. 1629:19-1630:8; see also DTX-98 1 (1997 academic publication observing “[a] full evaluation of possible variations in crystallography that might be encountered is now essential for the development for a new drug compound because the Food and Drug Administration (FDA) requires that analytical procedures be used to detect polymorphic, hydrated, or amorphous forms of the drug substance”).)
The Court adopts the '101 Patent's POSA definition for its '283 obviousness determination. (See Pl.'s Proposed Findings of Fact ¶ 1465 (“the person of ordinary skill in the art for purposes of the '283 Patent is the same as the POSA for the '101 Patent”); Defs.' Proposed Findings of Fact ¶ 1723 (“The '283 is invalid for obviousness whether the Court adopts Dr. Sacchetti's or Dr. Myerson's definition of a POSA.”).
At trial, Dr. Steed explained that prior art such as the J. Keith Guillory chapter from Harry G. Brittain's 1997 “Polymorphism in Pharmaceutical Solids, ” (DTX-12514), taught the POSA that the first step in conducting a polymorph screen is to use “solvents that are used in the processing of the drug substance in the first place. And so a [POSA] would be aware of the '052 [P]ublication. And the '052 [P]ublication is a roadmap basically that teaches that apremilast can be recrystallized from a binary solvent containing ethanol and acetone.” (Trial Tr. 1122:16-24.) As Dr. Steed testified, although “a polymorph screen is a campaign of experimental tests that can include hundreds or thousands of experiments, ” these tests are “very trivial experiments so robotic normal produced screens can easily run into the thousands.” (Trial Tr. 1155:9-15 (Steed).)
At trial, Zydus's expert also testified that fast cooling would inherently produce apremilast Form A crystals exhibiting the five peaks in claim 2 of the '283 Patent. (Trial Tr. 1190:24-1191:2 (Sacchetti).) Furthermore, Dr. Sacchetti testified that other prior art such as 1994 publications from Feiser, Guillory, and Byrn describe recrystallization by fast cooling. (Trial Tr. 1190:24-1191:2, 1188:23-1189:19 (Sacchetti); DTX-1013-6 (Byrn 1994); DTX-1259-16 (Guillory); JTX-178 5-9, (Fieser).) According to Dr. Sacchetti, based on the therapeutic utility of apremilast as described in the '052 Publication, a POSA would have been motivated to formulate enantiomerically pure apremilast, and would have tried the fast cooling method. (Sacchetti Tr. 1191:5-17.) Amgen's expert, Dr. Myerson, disputed Dr. Sacchetti's conclusion that Example 2 in the '052 Publication could produce apremilast Form A when combined with a fast cooling method. Dr. Myerson testified that the 2:1 acetone-ethanol ratio described in Example 2 would “clearly not” produce Form A, although, in the right ratio, particularly a 10:1 ethanol-acetone ratio, fast cooling could make Form A. (Trial Tr. 1614:9-20 (Myerson).) With respect to the '283 Patent claims, Dr. Myerson testified that in data from thirteen experiments submitted to the EPO reproducing the process outlined in Example 2, “if you're using a 2:1 ratio of ethanol to acetone, you use fast cooling or slow cooling, you always get Form B.” (Trial Tr. 1596:19-22.)
Nevertheless, as Zydus notes, Dr. Myerson conceded on cross-examination that a “POSA seeing [the] preclinical examples here in '052 would have been motivated to do a polymorph screen.” (Trial Tr. 1630:9-13 (Myerson).) In doing that screening, the POSA would have looked to “the only example for how to make enantiomerically pure apremilast” in the '052 Publication, Example 2. (Trial Tr. 1630:14-18 (Myerson).) Evidence at trial established that Guillory taught a POSA that in conducting a polymorph screen, “a screen should include solvents encountered during formulation and processing, ” meaning acetone and ethanol in the context of the '052 Publication. (Trial Tr. 1628:3-11.) Amgen's expert agreed that a POSA would have been able to design a polymorph screen using ethanol, acetone, and mixtures thereof, and varying cooling rates, including fast cooling. (Trial Tr. 1631:20-25 (Myerson).)
On the other hand, as the evidence at trial established, crystallization and polymorphism are highly unpredictable. (Trial Tr. 1609:16:23 (Myerson) (“At any date you wish to pick, 2002, 2008, or today” one “can't predict whether a compound has multiple crystalline forms” or “[t]he properties of those forms including the XRPD peaks, what. . . specific crystalline form, and how the many experimental variables impact whether a crystal will form, and if so, what specific crystalline form”). And although Zydus cites the '052 Publication in its obviousness combinations, that reference “teaches the 2:1 ratio of ethanol to acetone as the final crystallization step.” (Trial Tr. 1630:23-1631:3 (Myerson).) But as discussed above with respect to the' 101 Patent, the Court has already credited Dr. Myerson's testimony that “[a]s we saw in the EPO data, if you're using a 2:1 ratio of ethanol to acetone, you use fast cooling or slow cooling, you always get Form B.” (Trial Tr. 1596:19-22 (Myerson).)
Accordingly, as Dr. Myerson explained, even where a POSA was motivated to perform routine polymorph screens on the '052 Publication's Example 2 involving varying cooling rates or varying mixtures of ethanol and acetone, the POSA “can't have a reasonable expectation of success of knowing-of making something”-here Form A-that “[they] don't know exists.” (Trial Tr. 1609:7-8, 1608:23-1609:6 (Myerson) (“[E]xample 2 makes Form B. So Dr. Sacchetti's obviousness combination would not make Form A because his obvious combination is Example 2 from the '052 Publication and polymorph references. A POSA would have no reasonable expectation of success in making Form A based on the '052 Publication because polymorphism is very unpredictable and a POSA could have no reasonable expectation of making Form A because they don't know [it] exists nor what its properties are.”); see also PTX-1114260 (“Every compound is essentially a new situation, and the state of our knowledge and understanding of the phenomenon of polymorphism is still such that we cannot predict with any degree of confidence if a compound will be polymorphic, prescribe how to make possible (unknown) polymorphs, or predict what their properties might be.”).)
Because a polymorph screen of Example 2 of the '052 Publication would have had no reasonable expectation of success in producing Form A, a then-unknown crystal, Zydus's obviousness arguments fail. Bristol-Myers Co. v. U.S. Int'l Trade Comm'n, No. 89-1530, 1989 WL 147230, at *4 (Fed. Cir. Dec. 8, 1989) (“[A] new crystalline form of a compound would not have been obvious absent evidence that the prior art suggests the particular structure or form of the compound or composition as well as suitable methods of obtaining that structure or form.” (emphasis added) (citation omitted)). “‘[G]eneral motivation to discover an undefined solution that could take many possible forms' is insufficient to establish obviousness where the prior art does not suggest the unknown claimed form.'” Kowa Co., Ltd. v. Amneal Pharms., LLC, No. 14-7934, 2017 WL 10667089, at *28 (S.D.N.Y. Sept. 19, 2017) (quoting In re Armodafmil, 939 F.Supp.2d 456, 500 (D. Del. 2013)).
The Court's opinion in this regard is consistent with several district courts that have considered similar obviousness invalidity arguments involving polymorph screens. See, e.g., In re DepomedPatentLitig., 2016 WL 7163647, at *53 (D.N.J. Sept. 30, 2016) (“[T]he determination of the structure and properties of the polymorph forms of tapentadol hydrochloride-would have been impossible to predict. . . . Defendants have failed to meet their burden of demonstrating that the specific polymorph Form A of tapentadol hydrochloride was obvious.”); Kowa Co., Ltd. v. AmnealPharms., LLC, No. 14-7934, 2017 WL 10667089, at *26 (S.D.N.Y. Sept. 19, 2017) (“The unpredictable possible results a POSA could obtain from performing a polymorph screen of pitavastatin calcium are a far cry from evidence of a ‘finite number of identified, predictable solutions' that the Federal Circuit has declared ‘might support an inference of obviousness.'” (quoting Eisai Co. Ltd. v. Dr. Reddy's Labs., Ltd., 533 F.3d 1353, 1359 (Fed. Cir. 2008))); In re Armodafinil, 939 F.Supp.2d at 498 (“[T]here was no way to reasonably predict the outcome of any vast number of possible conditions that could have been chosen, as the selection of certain sets of conditions, but not others, could have resulted exclusively in forms other than Form I or mixtures of forms.”); Merck & Cie v. Watson Labs., Inc., 125 F.Supp.3d 503, 514 (D. Del. 2015) (“Both sides' experts agree that finding an unknown polymorph requires experimentation. While there may have been a motivation to discover new crystalline polymorphs of MTHF, doing so would have required a process of trial and error. There was therefore no reasonable expectation of success of finding Type I crystals.”), rev'd on other grounds, 822 F.3d 1347 (Fed. Cir. 2016).
2. Anticipation and the '283 Patent
Finally, the Court considers Zydus's invalidity challenge to the '283 Patent on anticipation grounds. “A prior art reference which expressly or inherently contains each and every limitation of the claimed subject matter anticipates and invalidates.” Schering Corp. v. Geneva Pharms., 339 F.3d 1373, 1379 (Fed. Cir. 2003). Although Zydus argues that Example 2 of the '052 Publication inherently results in Form A of apremilast when fast cooling is employed, as Amgen notes, Zydus did not present any experimental evidence that Example 2 of the '052 Publication ever produces Form A. On the contrary, Dr. Myerson testified that in data from thirteen experiments submitted to the EPO reproducing the process outlined in Example 2, “if you're using a 2:1 ratio of ethanol to acetone, you use fast cooling or slow cooling, you always get Form B.” (Trial Tr. 1596:19-22.) The Court, accordingly, finds that Zydus has failed to meet its burden of demonstrating that Example 2 of the '052 Publication anticipates claims 2 and 27 of the '283 Patent.
IV. CONCLUSION
For the reasons set forth above, the Court will enter judgment in favor of Amgen in part and for Defendants in part.