Most drugs are covered by multiple patents, with initial patents directed broadly to the compound and later patents directed to increasingly narrower uses of the compound. This provides opportunities for the compound to be approved as a generic drug before expiration of all of the patents, based on a “skinny” label – i.e., a label restricted to uses not covered by the later “use” patents. This strategy is contentious. It provides for competition but sometimes results in infringement due to off-label use of the generic drug – a delicate balance. In the case of H. Lundbeck A/S et al. v. Lupin et al., the Federal Circuit confirmed that an appropriately crafted skinny label cannot be the basis for a finding of infringement. Its ruling may encourage the strategy, but also provides insights for development and enforcement of use patents.

Takeda’s NDA for Trintellix® (vortioxetine)

Trintellix® (vortioxetine) is a drug for treating major depressive disorder in adults by enhancing serotonergic activity in the central nervous system. It was first approved by the U.S. FDA in 2013 under NDA No. 204447. Sales of Trintellix totaled $694 million in 2021. The drug was originally developed by Lundbeck and later licensed to Takeda, who is the holder of the NDA.

The active ingredient of Trintellix® is a salt of vortioxetine. The compound is covered by two patents: US7,144,884 (“the ’884 patent”) and US8,476,279 (“the ’279 patent). The ’884 patent claims the compound and a method of treating an affective disorder with the compound. The ’279 patent claims a pharmaceutical composition comprising the compound, and a method of treating an affective disorder with the pharmaceutical composition. The ’884 patent and the ’279 patent (“the compound patents”) have expiration dates of June 17, 2026, and October 2, 2022, respectively. In this case, Defendants’ ANDAs will only become effective after the expiration of the compound patents.

The Patents-at-Issue in the Litigation

This litigation involves three patents issued after the compound patents: US9,278,096 (“the ’096 patent”), US9,125,910 (“the ’910 patent”), and US9,101,626 (“the ’626 patent”).

The ’096 patent and ’910 patent (“the use patents”) cover methods of treating certain depression patient populations with vortioxetine, and have expiration dates of March 21, 2032, and June 15, 2027, respectively. The ’096 patent claims a method of treating depression, anxiety, or abuse or chronic pain, where the patient has previously received or is receiving another medication for treating such disease, but the medication has been ceased or reduced due to sexually related adverse events. The ’910 patent claims a method for treating cognitive impairment in a depression patient.

Defendants’ ANDAs seek approval of only one indication, namely, major depressive disorder (MDD) in adults. In contrast, the ’096 patent covers treating a sub-population of depression patients who could not tolerate a previous treatment and the ’910 patent covers treating depression patients with additional symptoms. The ’096 patent is based on two clinical studies demonstrating improved sexual functioning in patients when switching to vortioxetine from other drugs. The label of Trintellix® includes these studies in its “Clinical Studies” section. Here, Defendants’ ANDA labels carved out these data from the Trintellix®’s label, and instruct physicians not to compare rates of adverse reactions of sexual dysfunction between vortioxetine and other drugs.

The ’626 patent covers a process for manufacturing vortioxetine, and has an expiration date of June 15, 2027.

Infringement of the Use Patents

As discussed above, the Orange Book patents for Trintellix® (vortioxetine) include patents on the drug compound and its use for treating major depressive disorder (MDD) that are expired or that will expire before any generic drug would be approved. Later-expiring patents include claims that are also directed to the use of Trintellix® for treating MDD (among other things), but which are limited to situations where the patient has been previously treated and had adverse reactions involving either sexual or cognitive capabilities.

The plaintiff argued that if the ANDA filer were to market vortioxetine for treatment of MDD, as proposed, it would infringe the later-expiring patents. Specifically, it argued that the ANDA filer would infringe claim 7 of the ’096 Patent and claim 6 of the ’910 Patent on three grounds: (1) under the plain text of 35 U.S.C. § 271(e)(2)(A) based on the ANDA, (2) indirectly by inducing infringement under 35 U.S.C. § 271(b), and (3) indirectly as a contributory infringer under 35 U.S.C. § 271(c).

As noted above, the claims at issue covered a subset of the treatments covered by the initial patents and the ANDA. All of the plaintiffs arguments for infringement assumed that the treatments covered by the initial patents and the ANDA would, in fact, encompass those covered by the claims at issue. The Court appears to have acknowledged as much, stating, “[s]ome drugs for the treatment of depression are associated with high rate of sexual dysfunction” and “[c]ognitive impairment is very common in MDD patients.”

The Court nonetheless swiftly rejected the first basis, pointing to the language of the statute, which establishes infringement based on the filing of an ANDA “for a drug … the use of which is claimed in a patent.” As stated by the Court, the statute requires that the infringing use “must be the use for which an applicant is seeking marketing approval.” In other words, the use to be approved by the ANDA must be the same as the use claimed in the patent to have an independent basis of infringement under 35 U.S.C. § 271(e)(2)(A).

The Court rejected the second basis, noting that the allegations of induced infringement “relied solely on defendants’ proposed ANDA labels as the inducing conduct” — with no evidence of advertising or promotional materials that might otherwise support allegations of induced infringement. The Court found that where the label is the sole basis for the alleged inducement of infringement, there can only be infringement where the use advocated in the label is the same as the patented use.

The Court made this point emphatically, as a matter of principle, stating:

[I]t cannot be, as plaintiffs suggest, that a patentee can bar the sale of a drug for a use covered only by patents that will have expired simply by securing a new patent for an additional, narrower use ….

[W]e do not see how, in the normal course, a label required to market the drug for a use covered by expired patents could demonstrate the required specific intent to encourage infringement of new patents covering different uses.

The Court also rejected any suggestion that an expectation that practice of the earlier patents would include practice of the later patents, stating “mere knowledge of possible infringement by others does not amount to inducement; specific intent and action to induce infringement must be proven.”

The Court rejected the third basis, concluding there was no contributory infringement, based on the District Court’s finding that there would be substantial noninfringing uses of vortioxetine. Specifically, regarding the ’096 Patent, noninfringing uses would include “prescribing vortioxetine to patients with no prior treatment, patients with prior treatment other than with the drugs referenced in the ’096 patent, and in cases where the prior antidepressant was ceased or reduced for reasons other than sexually related adverse events (for example, due to poor efficacy).” Regarding the ’910 patent, noninfringing uses would include “prescription for treating MDD, prescription to patients without cognitive impairment, and prescription for purposes unrelated to cognition.”

Infringement of the Process of Preparation Patent

On a different issue, the district court had found that Lupin infringed the ’626 patent. In a cross-appeal, Lupin disagreed with the district court’s construction of the term “reacting” in a step of “reacting compound II…with a compound [III]…and a compound [IV]” recited in claim 12 of the ’626 patent. Adopting Plaintiff’s construction, the district court held that “reacting” means “the changing of a reactant(s) to product(s).”

In its appeal, Lupin argued that “reacting” means the “the specified chemicals are added to the reaction vessel at the beginning of the process as starting material,” and that Lupin’s process does not infringe because it does not use Compound II as a starting material. Lupin asserted that the prosecution history supports its construction because the examiner used the phrase “starting compound” to refer to the materials reacted together.

The Federal Circuit rejected this argument, stating that the examiner said nothing about the scope of “reacting,” and the patentee made no statement during prosecution that limited the scope of “reacting.” Further, pointing to dependent claim 3, which recites an addition limitation requiring the reactant compounds are mixed at the start of the process, the Court argued that Lupin’s construction is in tension with this claim and would render the claim superfluous. The Court agreed with the district court’s construction and its finding of infringement of the ’626 patent by Lupin’s process.


The Federal Circuit’s opinion clarifies and confirms that the filing of a Paragraph IV certification pursuant to 21 U.S.C. § 355(j)(2)(A)(vii)(IV) “gives the brand manufacturer a right to sue the ANDA filer for infringement,” but infringement is then determined “by traditional patent infringement analysis” under 35 U.S.C. § 271(a). Insofar as a plaintiff alleges indirect infringement and relies upon proposed labeling in the defendant’s ANDA, that label must correspond to the claims of the patent. Otherwise, the plaintiff must provide additional evidence to support a finding of indirect infringement.

This case also demonstrates the value of patents covering a drug’s manufacturing process. Although not listable on the Orange Book, these patents can still provide valuable protection for the brand drug. Generic companies should be aware that such patents can be brought by an NDA holder into a Hatch Waxman litigation like this case, and should take this into account when accessing infringement risk before launching a generic product.