ANDA Update

Stability of Ready-to-Use Dexmedetomidine Not Inherent

After a bench trial, the US District Court for the District of Delaware found several Hospira patents covering a diluted, “ready-to-use” composition of dexmedetomidine obvious in view of the concentrated prior art product—but found that a claim requiring dexmedetomidine to be stable was not inherent or obvious, and was infringed by Amneal’s abbreviated new drug application (ANDA) submission as a matter of law. Hospira, Inc. v. Amneal Pharmaceuticals, Case No. 285 F.Supp.3d 776 (D. Del.,Jan. 22, 2018) (Andrews, J).

In December 1999, Hospira obtained US Food and Drug Administration (FDA) approval for its original Precedex product containing the sedative dexmedetomidine. Precedex was and is sold in a 2 mL glass vial containing 100 μg/mL of dexmedetomidine hydrochloride. Before being administered, Precedex concentrate must be diluted to an appropriate concentration per instructions on the Precedex label.

Hospira owned a compound patent covering dexmedetomidine that expired in 2013. Shortly before its compound patent expired, Hospira received approval for a “ready-to-use” formulation of dexmedetomidine (Precedex premix) containing 4 μg/mL of dexmedetomidine in 20, 50 or 100 mL vials. Hospira obtained several patents covering the Precedex premix formulation and methods of using it. These patents included limitations related to specific concentrations of a “ready-to-use” formulation of various volumes stored in a sealed glass container.

In 2015, Amneal filed an ANDA seeking FDA approval for a generic version of Precedex premix. Hospira filed suit alleging that the approved product infringed Hospira’s patents.

At trial, Amneal argued that the patents were obvious in view of the original Precedex product and label, and the knowledge of one of skill in the art. Hospira admitted that many of the claim limitations (including those related to dexmedetomidine concentration and total volume) would be met by diluting the original Precedex product in accordance with the instructions on its label. However, Hospira argued that one of ordinary skill would not be motivated to prepare a “ready-to-use” formulation and would not have a reasonable expectation of success in using a sealed glass container to store the formulations.

The court disagreed. Specifically, the court pointed to prior art articles showing that physicians had a preference for ready-to-use formulations, and that one hospital had even created its own ready-to-use Precedex syringes. In addition, the court noted that glass was the go-to material for containers for small-volume injections, and that the original Precedex product was sold in a glass vial. Thus, the court found all but one of the asserted patents obvious.

The remaining patent required that the formulation exhibit no more than a 2 percent decrease in dexmedetomidine concentration after storage for five months. Amneal argued that this limitation was obvious and pointed to testing of Precedex premix to show it was stable for five months. However, the court found that because Amneal did not provide any scientific explanation confirming the inherency, the examples Amneal pointed to were not enough. Thus, the remaining patent was found valid.

The court next turned to infringement. The only disputed question was whether Amneal’s product exhibited the required stability. Hospira argued that Amneal infringed as a matter of law because Amneal’s ANDA specified that the approved product’s dexmedetomidine concentration would remain within 90 percent of its original concentration for the entirety of its two-year shelf life. The court agreed. Since approval of Amneal’s ANDA (allowing Amneal to manufacture a product exhibiting no more than a 10 percent decrease in concentration after two years) would allow Amneal to manufacture a product exhibiting no more than a 2 percent decrease in concentration after five months, Amneal infringed as a matter of law. This was true even though the court found that as a matter of fact, Hospira did not prove that the stability data submitted for Amneal’s product showed that the product actually exhibited no more than a 2 percent decrease in concentration after five months.

How Many Infringers Does It Take to “Administer” a Medication?

Addressing the issue of whether defendant’s abbreviated new drug application (ANDA) formulation would infringe plaintiff’s method patents under a theory of joint direct infringement, Circuit Judge Bryson sitting in the US District Court for the District of Delaware denied plaintiff’s motion for summary judgment. Pernix Ireland Pain DAC v. Alvogen Malta Operations Ltd., Civil Action No. 16-139-WCB, 2018 (D. Del., May 15, 2018) (Bryson, J).

Pernix filed a motion for summary judgment of infringement of its method patents based on Alvogen’s filing of an ANDA with the US Food and Drug Administration (FDA) seeking authorization to sell hydrocodone bitartrate extended release capsules as generic versions of Pernix’s formulation, sold under the name Zohydro.

During claim construction, the court rejected Pernix’s argument that the term “administering” should include the acts of prescribing or dispensing by a physician or health care provider, and issued a construction that “administering” meant “delivering into the body.” In the context of this construction, Pernix filed an infringement contention based on a theory of joint direct infringement by the physician and the patient. Specifically, Pernix contended that a physician and patient jointly infringe the method claims when the physician prescribes and directs a hepatically impaired patient to take extended release hydrocodone bitartrate, and the patient takes the drug as directed.

The court cited to established Federal Circuit precedent that “joint direct infringement of a claimed method occurs if the physician ‘directs or controls’ the patient so that the patient’s acts are attributable to the physician.” Eli Lilly & Co. v. Teva Parenteral Meds., Inc., Case No. 845 F.3d 1357, 1364 (Fed. Cir., 2017). In particular, the court noted that the requisite degree of direction and control occurs if the physician “conditions participation in an activity or receipt of a benefit” on the patient’s performance of one or more steps of the claimed method, and the physician “establishes the manner or timing of that performance.”

The court found that “[a]lthough it is clear that a physician gives instructions to a patient as to dosing levels and practices, the extent to which those instructions constitute ‘direction and control’ of the patient’s infringing conduct turns on factual questions.” Pernix argued that no issue of material fact existed regarding a physician’s direction and control. Specifically, Pernix cited to the fact that hydrocodone is a controlled substance, and physicians enter into agreements with their patients requiring the patients to take the drug as prescribed, and thus exercise control and direction over the patients’ conduct in that regard. Pernix also cited to the deposition testimony of Alvogen’s infringement expert that physicians establish the timing and manner in which the patients administer the drug.

Alvogen countered that Pernix’s evidence did not show that physicians necessarily control and direct patients, because there is no evidence that physicians “condition” on the patient’s administering the drug as prescribed. In addition, Alvogen argued that physicians have no means of verifying that instructions are followed.

The court found that Pernix’s evidence was insufficient to show that there was not an issue of material fact, stating that “[w]hile Pernix’s evidence could support a finding that physicians exercise direction and control over the self-administration of hydrocodone extended release formulations by patients, the evidence on that issue is not so compelling as to justify the entry of summary judgment in Pernix’s favor.”

Practice Note: Alvogen was able to provide expert testimony that countered the evidence proffered by Pernix. This expert testimony was enough to show that the facts regarding joint direct infringement were disputable, thus demonstrating the value of expert testimony at the summary judgment stage.

Teva’s Skinny Label Found Not to Induce Infringement by Off-Label Use

The US District Court for the District of Delaware granted in part Teva Pharmaceuticals’ renewed motion for judgment as a matter of law (JMOL), finding that substantial evidence did not support the jury’s finding of induced infringement, and thus the jury’s award of more than $235 million could not stand. GlaxoSmithKline LLC v. Teva Pharm. USA, Inc., Case No. CV 14-878-LPS-CJB (D. Del., Mar. 8, 2018) (Stark, J).

The case involved GlaxoSmithKline’s (GSK’s) branded version of the drug carvedilol. GSK had received a patent directed towards methods for decreasing congestive heart failure (CHF) mortality by administering carvedilol. GSK also received US Food and Drug Administration (FDA) approval to market its drug under the brand name Coreg for three indications: hypertension, CHF and post-heart-attack ventricular dysfunction. Despite receiving approval for three indications, GSK marketed the drug only for CHF.

In 2007, Teva filed an abbreviated new drug application along with a “Section viii statement,” also known as a skinny label or a carve out, which allows the generic to obtain approval to sell the drug as long as it excludes from the label methods of use that are under patent. Here, the FDA approved Teva’s generic carvedilol drug with its skinny label, which excluded the indication for CHF. Several years later, Teva revised its label to include the CHF indication once GSK’s patent covering methods for decreasing CHF was de-listed from the Orange Book.

At issue was whether substantial evidence supported the jury’s finding that Teva induced infringement of the claims in GSK’s patent. Teva’s skinny label did not instruct physicians to prescribe generic carvedilol to treat CHF. The only approved indications on the skinny label were for hypertension and post-heart-attack ventricular dysfunction.

GSK presented evidence that at least one physician had prescribed Teva’s generic version of carvedilol “off-label” for the treatment of CHF in a patient. The court, however, found this insufficient to demonstrate that Teva’s alleged inducement caused doctors to infringe GSK’s patents. Not only did the label not instruct physicians to infringe GSK patents, but physicians who prescribed generic carvedilol for off-label uses relied on sources other than Teva’s label and marketing materials, the court found. Such physicians used their own experience and had access to medical journals, guidelines and GSK’s promotional materials that discussed the benefits of carvedilol.

Notably, GSK’s own expert witness testified that he had not read Teva’s generic label before he started writing prescriptions for carvedilol. GSK also conceded that physicians knew how to use carvedilol for treating CHF prior to the entrance of Teva’s generic carvedilol in the market.

The court found that GSK presented no direct evidence that Teva induced any doctor to prescribe generic carvedilol to treat CHF. Therefore, the court granted Teva’s JMOL on inducement of infringement and willful infringement.

Practice Note: The court’s finding that a generic drug maker cannot induce infringement of a method of use patent for an indication that is not even listed on its drug label seems intuitive. The court’s decision, however, highlights an obstacle that branded companies face in enforcing their patents. Even though a generic drug maker may exclude the indications covered by a patent, physicians may still prescribe the generic for the “off-label” patented use based on their broader understanding of the drug. In order for generic drug manufacturers to be found liable for induced infringement, however, the patent holder must prove that the actions of the generic company actually caused the physicians to directly infringe the patent. This may not always be the case—for example, here, GSK’s expert did not even read the generic label before prescribing Teva’s carvedilol.

Lack of Efficacy Data in Prior Art Defeats Invalidity Challenges

The US Court of Appeals for the Federal Circuit affirmed a district court decision that the asserted claims directed towards the anti-epileptic drug lacosamide were not invalid based on obviousness-type double patenting, obviousness or anticipation. UCB, Inc. v. Accord Healthcare, Inc., Case No. 2018 (Fed. Cir., May 23, 2018) (Stoll, J) (Prost, CJ, dissenting).

UCB owns a patent disclosing and claiming lacosamide, RE 38,551. Lacosamide belongs to a class of compounds known as functional amino acids (FAAs). In particular, lacosamide is an FAA with three chemical groups at positions R, R1 and R3. R is a benzyl group, R1 is a methyl group, and R3 is a nonaromatic methoxymethyl group. Lacosamide also contains only the R enantiomer.

UCB also holds an approved new drug application covering lacosamide, marketed as Vimpat, for the treatment of epilepsy. Several generic drug makers (appellants) filed an abbreviated new drug application seeking to market a generic version of lacosamide along with a Paragraph IV certification stating that the ’551 patent is invalid or unenforceable, or that their proposed product will not infringe upon the ’551 patent.

Consequently, UCB sued appellants for patent infringement. Appellants stipulated to infringement of the ’551 patent but raised three affirmative invalidity defenses:

• Obviousness-type double patenting
• Obviousness
• Anticipation

The district court determined that the claims were not invalid based on these ground. The generic drug manufacturers appealed.

Appellants first argued that the asserted claims of the’551 patent were not patentably distinct from claims 44–47 of US Patent No. 5,654,301, which discloses a genus of chemical compounds and recites several different groups that can be placed at the R and R1 positions. The ’301 patent does not, however, specifically disclose the chemical structure of lacosamide.

The Federal Circuit affirmed the district court’s finding that the ’551 patent was patentably distinct from the claims of the ’301 patent. Unlike the claims of the ’301 patent, the ’551 patent requires (1) the R enantiomer with 90 percent purity, (2) an unsubstituted benzyl at R and (3) an unsubstituted methyl at R1. The Court agreed that “the trial evidence supports the district court’s finding that there was no prior art that would have provided a person of ordinary skill reason to believe that unsubstituted benzyl and methyl would have been successful.”

Chief Judge Prost’s dissent faulted the district court for not considering LeGall in its primary double-patenting analysis. LeGall discloses a compound, 107e, which is the racemate version of the lacosamide compound—a mixture of both the R and S enantiomers. While LeGall does not provide any data regarding the anticonvulsant efficacy of 107e, it speculates that based on its structure, 107e “may have good anticonvulsant activity.” The majority, however, disagreed, pointing to expert testimony at trial and to the district court’s explanation that LeGall’s reference to compound 107e lacked any data or discussion that would have motivated a person of ordinary skill in the art to use a nonaromatic compound such as a methoxymethyl at R3.

Appellants also argued that US Patent No. 5,378,729’s disclosure of a genus of FAAs that would encompass lacosamide, as well as its disclosures that benzyl is “especially preferred” and that methyl is “most preferred” for this genus of compounds, itself renders the asserted claims obvious. The Court disagreed, finding that this disclosure did not provide any indication that out of the millions of possible choices, an unsubstituted benzyl at R and an unsubstituted methyl at R1 would have been selected in combination with a methoxymethyl group at R3 to arrive at lacosamide.

Next, the court addressed appellant’s obviousness arguments. Appellants had asserted that claim 9 of the ’551 patent would have been obvious based on LeGall alone or in combination with the ’729 patent and Kohn 1991. The district court concluded that a person of ordinary skill in the art would not have selected any FAA, let alone the compounds disclosed in LeGall and Kohn 1991, as a lead compound. The Court agreed with the district court’s finding that none of the prior art data cited showed any efficacy for lacosamide as an anticonvulsant. The class of compounds tested in LeGall showed little potency, and the compound disclosed in Kohn 1991 proved to be relatively unstable.

Finally, the Court affirmed the district court’s finding that the racemate version of lacosamide (107e) disclosed in LeGall did not anticipate the asserted claims of the ’551 patent directed towards the R enantiomer lacosamide, because the reference does not disclose separation into individual enantiomers, nor does it disclose any pharmaceutical data of the R enantiomer recited in claim 9 of the ’551 patent.

Practice Note: Lack of relevant efficacy data and expert testimony that there was no reasonable expectation of success may help a patent survive obviousness type double patenting and obviousness invalidity challenges to species claims based on prior art patents disclosing a genus of chemical compounds.

Claim Scope Must Include the Recited Compound

The US Court of Appeals for the Federal Circuit affirmed a lower court ruling that a claim must at a minimum include the recited enantiomer. Sumitomo Dainippon Pharma Co., Ltd. v. Emcure Pharmaceuticals Ltd., et al., Case No. 887 F.3d 1153 (Fed. Cir., Apr. 16, 2018) (Stoll, J).

The case arose under the Hatch-Waxman Act and involved a patent covering lurasidone, the active ingredient in the schizophrenia and bipolar depression drug LATUDA. Appellants sought approval to market generic versions of LATUDA. Plaintiffs initiated the lawsuit and asserted claim 14 of the patent, which recites an imide compound of a particular formula or an acid additional salt thereof.

The parties agreed that the specific formula depicted in the claim is lurasidone, a (-) enantiomer. Enantiomers can be of the (-) or (+) type, and a mixture with 50 percent of each of the (-) and (+) enantiomer is known as a racemate or racemic mixture. Appellants sought to limit claim 14 to “racemic mixture of two enantiomers of which the structural formula is representative.” The district court rejected appellants’ construction, explaining that such a construction would exclude the (-) enantiomer depicted in claim 14. Thereafter, the district court construed the claim as covering “lurasidone, lurasidone’s enantiomer as well as mixtures of these enantiomers.” Appellants appealed.

On appeal, the Federal Circuit affirmed the district’s court finding that the claim covered at a minimum the (-) enantiomer. The Federal Circuit, however, declined to address the rest of the lower court’s construction, explaining that appellants had conceded infringement under such a construction.

The Federal Circuit began its analysis by examining the claim language. The Court noted that appellants did not dispute that a skilled artisan “looking at claim 14’s structure in a vacuum would understand it to be one way of depicting the (-) enantiomer.” By contrast, nothing in the claim language limited the claim scope to a racemate, the Court explained.

Turning next to the patent specification, the Court noted that the specification describes the enantiomer as a preferred embodiment and does not disclaim the (-) enantiomer. The specification does not disparage the (-) enantiomer, nor does it refer to the racemic mixture as forming the basis for the invention or all the patent’s embodiments. As for appellants’ argument that the claimed structure was the same as compound 101, which appellants contended was a racemate, the Court disagreed that compound 101 was a racemic mixture and further concluded that the specification does not define claim 14’s structure as compound 101.

The Court also did not find appellants’ extrinsic evidence in the form of textbooks and expert testimony compelling. Noting that extrinsic evidence is in general “less significant than intrinsic evidence,” the Court explained that even though the appellants’ expert had argued that it was conventional in the art to use a single enantiomer to depict a racemate, the expert did not state that a person of ordinary skill in the art would understand a “depiction of a single enantiomer to exclude the very enantiomer depicted.”

Practice Note: A claim construction that excludes the recited structure is likely to be rejected.

Throwing Metaphorical Darts: When Is Obvious to Try Not So Obvious?

Addressing whether a patented claim for a formulation was obvious to try, the district court on a motion for summary judgment held that the claim was nonobvious where several prior art references resulted in numerous approaches to a problem and was thus less “obvious to try.” Valeant Pharm. Int’l Inc. v. Mylan Pharm. Inc., Case No. 15-8180, 2018 (D.N.J., May 1, 2018) (Chesler, J).

Valeant Pharmaceuticals International, Inc., and Wyeth LLC (collectively, plaintiffs) owned US Patent No. 8,552,025, which covers Valeant’s drug Relistor (methylnaltrexone). Several generic drug makers (defendants) sought to make and sell a generic version of Relistor prior to the expiration of the relevant patents, and this Hatch-Waxman suit was initiated.

At issue in this motion for partial summary judgment was claim 8 of the ’025 patent. The disputed claim element is a pharmaceutical parenteral formulation of methylnaltrexone in saline at a pH range of 3.0–4.0 with 24 months stability at ambient conditions. The plaintiffs moved for partial summary judgment of no invalidity. The defendants asserted that the claimed pH range of 3.0–4.0 was obvious to try due to prior art with an overlapping pH range of 3–7 for solutions of naloxone and naltrexone. The defendants also introduced numerous prior art references to prove that adjusting the pH parameter of a formulation would have been obvious to try, since pH is one of many commonly adjusted formulation parameters to enhance stability, and prolonged stability is typically desirable.

In response, the plaintiffs relied on the prior art’s failure to teach that an acidic pH alone was responsible for improving the stability of a formulation using a naloxone or naltrexone. In the plaintiffs’ view, there was no motivation to combine prior art, which specifically taught that stabilizers or chelating agents, rather than pH, improved the stability of acidic solution formulations of naloxone derivatives. Moreover, none of the prior art formulations resulted in stability out to 24 months.

The court granted plaintiffs’ motion for partial summary judgment that the patent was not invalid as obvious because using the claimed pH range of 3.0–4.0 to achieve 24 months stability in light of prior art was unpredictable. The court ruled that the overlapping pH ranges were not obvious because the claimed pH range applies to methylnaltrexone, a quaternary amine, while the prior art pH range applies to naloxone and naltrexone, both tertiary amines that are different molecules from the patented claim. Also, achieving 24 months stability at ambient conditions by adjusting pH alone was not a predictable result because the prior art provided many approaches to improve formulation stability in addition to pH, including stabilizers, antioxidants, chelating agents, container closure and preservatives, without specific guidance to adjust pH. None of the prior art taught adjusting a formulation’s pH to improve stability, but rather recommended the use of chelating agents.

Practice Note: To defeat summary judgment on obviousness, raise a dispute of fact using secondary considerations, since these are fact based. The court must consider secondary considerations in its obviousness analysis. Also, asserting too many prior art references may result in cumulatively more options for an inventor to pursue and would thus make the claimed invention nonobvious.

*Candace Polster, summer associate in McDermott’s Chicago office, contributed this article.

Second Circuit’s Mandate Rule Limits New Causation Theories on Remand

On remand, indirect purchasers of the drug ACTOS sought to amend their antitrust complaint to modify their causal theory of harm. The district court denied plaintiffs’ first set of amendments under the mandate rule, which prohibits re-litigation of issues decided by an appellate court in the same case. However, the court permitted a second set of amendments that may have the same practical effect as the first. In re Actos End-Payor Antitrust Litig., Case No. 13-9244, 2018 (S.D.N.Y., Feb. 12, 2018).

In 1999, Takeda gained US Food and Drug Administration (FDA) approval to sell the diabetes drug ACTOS, which is covered by three patents. In the FDA’s Orange Book, Takeda listed its 4,687,777 patent as a drug-product patent, and its 5,965,584 and 6,329,404 patents as method-of-use patents. Takeda also allegedly made non-public statements to the FDA in 1999 and 2002 that the ‘584 and ‘404 patents were method-of-use and drug-product patents.

Starting in 2003, several generics began the process of applying to enter the ACTOS market upon expiration of the ’777 patent. With their abbreviated new drug applications (ANDAs), they submitted two types of certifications: (1) a Paragraph IV certification, which states that the branded company’s patents are “invalid or will not be infringed” by the generic company’s drug, and (2) a “Section viii statement,” which is a specific mechanism by which a generic may seek to market only new methods of using a particular drug, by “carving out” the methods of use covered by the branded company’s patent. Submitting an ANDA application with a Section viii statement presents a lower risk of a patent infringement lawsuit. Most of the generic companies filed both the Paragraph IV certification and a Section viii statement, but Teva filed only a Section viii statement.

Takeda sued the Paragraph IV filers in 2003 for patent infringement and six years later sued Teva. In 2010, the FDA received a citizen petition asking the FDA to deny Teva’s ANDA for failing to file a Paragraph IV certification. Takeda (again) informed the FDA that its ‘584 and ‘404 patents were drug-product and method-of-use patents. Relying on that representation, the FDA granted the citizen petition and required all ANDAs, including Teva, to file Paragraph IV certifications for the ‘584 and ‘404 patents. Takeda then settled its infringement claims, allowing the first-filers and Teva to begin marketing generics four years before the ‘584 and ‘404 patents expired.

Indirect purchasers of ACTOS brought state antitrust claims alleging that Takeda delayed generic entry by filing false patent descriptions and making false statements to the FDA. The court dismissed the complaint for failing to plausibly allege a causal theory of harm. Plaintiffs appealed, relying on their two causation theories:

• That Takeda’s false descriptions and false statements in 1999 and 2002 required generics to file Paragraph IV certifications, which delayed generic entry
• That Takeda’s misrepresentations caused the FDA to grant the citizen petition in 2010, which required Teva to file a Paragraph IV certification and delayed its entry

On remand, plaintiffs sought to amend their complaint in two ways. First, even if the generics did not originally know of Takeda’s false patent descriptions, in the absence of these false statements the FDA would have required the generics to amend their ANDAs to address only methods of use. The generics would have then learned of Takeda’s false descriptions, filed only Section viii statements and entered earlier than they did. The district court denied this amendment under the “mandate rule.” The Second Circuit had created the “law of the case” by ruling that the generics would not likely have known about Takeda’s allegedly false descriptions, and the district court was bound by that ruling. These amendments were thus precluded by the mandate rule.

Second, plaintiffs sought to add to their theory that Takeda’s 2010 statements on the citizen petition delayed not only Teva’s entry, but other generics’ as well. The district court found this “a closer question.” The Second Circuit had found this theory plausible, at least as to Teva, because it did not depend on Teva’s knowledge of Takeda’s statements; instead, it relied on the FDA’s ruling on the citizen petition, which itself relied on Takeda’s statements. The district court also held that plaintiffs could add other generics’ delayed entry to this theory. The generics would have learned of the FDA’s ruling because such rulings are public and the generics would have been following the matter closely. These amendments were not barred by the mandate rule.

Ultimately, plaintiffs were allowed to allege “that Takeda’s misrepresentations caused the FDA’s ruling and that the FDA’s ruling in turn caused a delay in the generic’s entry.” Thus, the mandate rule may have foreclosed a route, but it did not foreclose the destination.