Biosimilars & the BPCIA Patent Dance

What is the BPCIA and how is it different from Hatch-Waxman?

The Biologics Price Competition and Innovation Act (BPCIA), enacted in 2010 as part of the Affordable Care Act, created an abbreviated FDA approval pathway for biosimilars — follow-on versions of biologic drugs — and a unique system for resolving the patent disputes between biosimilar applicants and the makers of the original 'reference product.' It is the biologics counterpart to the Hatch-Waxman Act, which since 1984 has governed small-molecule generic drugs, but the two regimes differ fundamentally because the products differ fundamentally. Small-molecule drugs (Hatch-Waxman) are chemically synthesized and can be copied identically — a generic is the 'same' active ingredient. Biologics (BPCIA) are large, complex proteins (antibodies, hormones, enzymes) produced in living cells; they cannot be copied exactly, so a follow-on is 'biosimilar' (highly similar with no clinically meaningful differences) rather than identical, and a higher 'interchangeable' designation requires additional showing. Key structural differences: (1) Application: ANDA/505(b)(2) under Hatch-Waxman vs the 351(k) biosimilar application under BPCIA. (2) Data exclusivity: 5 years for a new chemical entity under Hatch-Waxman vs 12 years from reference product licensure under BPCIA — the longest data exclusivity in US drug law. (3) Patent disclosure: Hatch-Waxman uses the public Orange Book patent listing and Paragraph IV certifications; the BPCIA uses a PRIVATE information exchange — the 'patent dance' — with no public list of the patents at issue. (4) Litigation trigger: a Paragraph IV certification is itself an act of infringement under Hatch-Waxman; the BPCIA instead runs patent disputes through the structured § 262(l) exchange. (5) Stay: Hatch-Waxman imposes an automatic 30-month stay on FDA approval; the BPCIA has no automatic stay.

What is the BPCIA 'patent dance'?

The 'patent dance' is the nickname for the structured, multi-step pre-litigation information exchange set out in 42 U.S.C. § 262(l), designed to identify and narrow the patent disputes between a biosimilar applicant and the reference product sponsor (RPS) before litigation. The choreography: (1) § 262(l)(2): within 20 days of FDA accepting the biosimilar's 351(k) application, the applicant discloses its application and manufacturing information to the RPS. (2) § 262(l)(3)(A): within 60 days, the RPS lists patents it could reasonably assert. (3) § 262(l)(3)(B)-(C): the biosimilar applicant gives detailed non-infringement/invalidity/unenforceability positions on each patent, and the RPS replies with its infringement positions. (4) § 262(l)(4)-(5): the parties negotiate which patents go to immediate litigation; if they can't agree, a statutory exchange sets the number and identity. (5) § 262(l)(6): the RPS files suit on those patents within 30 days (first phase). (6) § 262(l)(8): the biosimilar applicant gives 180 days' notice before commercial marketing, opening a window for second-phase patent litigation and possible preliminary injunction. The dance was intended to force early, complete information sharing so disputes resolve before launch. In practice it has been criticized as complex and gameable, and after the Supreme Court made the first step non-compellable (Sandoz v. Amgen, 2017), many biosimilar applicants now weigh whether to dance at all.

Is the patent dance mandatory after Sandoz v. Amgen?

No — the patent dance is effectively optional for the biosimilar applicant, which is the central holding of Sandoz Inc. v. Amgen Inc. (U.S. 2017). The Supreme Court addressed two questions: (1) Can the reference product sponsor force a biosimilar applicant to engage in the dance — specifically the § 262(l)(2)(A) disclosure of its application and manufacturing information — through a federal injunction? The Court held NO. The BPCIA specifies its own consequence for skipping the disclosure: under § 262(l)(9)(C) and § 271(e)(2)(C)(ii), if the applicant does not provide the application and manufacturing information, the RPS (but not the applicant) may immediately bring a declaratory judgment action for patent infringement, and the applicant loses the ability to control the timing and sequencing of litigation that the dance otherwise provides. That statutory remedy is exclusive — no federal injunction compelling the disclosure is available (the Court left open whether a state-law injunction might be, on remand). So an applicant can decline the dance and accept that the RPS can sue immediately on whatever patents it chooses. (2) Must the 180-day notice of commercial marketing (§ 262(l)(8)(A)) wait until after FDA licensure? The Court held NO — the notice is effective whether given before or after the FDA licenses the biosimilar, rejecting the Federal Circuit's rule that had effectively added six months of exclusivity by requiring post-licensure notice. Net effect: the dance's first step cannot be compelled, but skipping it carries real strategic cost (loss of litigation sequencing control and exposure to immediate DJ suit), while the 180-day commercial marketing notice remains a genuine, enforceable requirement that can be satisfied earlier than the Federal Circuit had allowed.

What is the 12-year exclusivity for biologics?

The BPCIA grants the reference biologic 12 years of regulatory data exclusivity from the date the reference product was first licensed by the FDA. Mechanically: (1) A biosimilar 351(k) application cannot be APPROVED by the FDA until 12 years after the reference product's initial licensure; and (2) a 351(k) application cannot even be SUBMITTED until 4 years after the reference product's licensure. This is distinct from patent protection — it is a regulatory bar that runs regardless of patent status, so it can outlast or undercut the reference product's patents depending on timing. The 12-year period is the longest data exclusivity in US pharmaceutical law, far exceeding the 5 years for small-molecule new chemical entities under Hatch-Waxman. Important nuances: (1) Exclusivity attaches to the reference product, and the statute restricts a separate 12-year period for subsequent related products from the same sponsor (to prevent 'evergreening' exclusivity by minor modifications) — a supplement or a change in route/dosage/strength to an already-licensed product does not generally restart the clock. (2) Pediatric studies can add 6 months (pediatric exclusivity) on top. (3) The 12-year figure has been a recurring policy target — proposals to shorten it (e.g., to 7 years) to accelerate biosimilar competition have been debated repeatedly but the 12-year period remains in effect. (4) Exclusivity and the patent dance are independent tracks: a biosimilar may clear all patents yet still wait out the 12-year exclusivity, or vice versa. For reference: the first FDA-approved biosimilar in the US was Zarxio (filgrastim-sndz) in 2015; the biosimilar market has since expanded substantially, particularly after biosimilars of major antibody products began launching.

Why are there no biologic patents in the Orange Book?

Because the Orange Book is a creature of the Hatch-Waxman small-molecule system, and biologics live under the separate BPCIA regime, which deliberately chose a private patent-exchange model instead of public listing. Under Hatch-Waxman, a brand drug company must list in the FDA's Orange Book the patents covering its approved small-molecule drug; a generic applicant then certifies against each listed patent (the Paragraph IV certification being the litigation trigger). The system is public and patent-by-patent transparent. The BPCIA took the opposite approach: there is no requirement to publicly list the patents covering a biologic, and the patents potentially in dispute are surfaced only privately, between the parties, during the § 262(l) patent dance — the RPS's patent list goes to the biosimilar applicant, not to a public registry. The FDA's 'Purple Book' is the biologics analogue to the Orange Book but it lists licensed biological products and their exclusivity periods — NOT the patents that could be asserted in litigation (though it has been expanded over time to include some patent information disclosed under certain provisions). Practical consequences: (1) Biosimilar developers cannot simply read a public list to map their patent risk — risk assessment depends on independent freedom-to-operate analysis and, ultimately, on what emerges in the dance. (2) The private exchange, combined with manufacturing-process patents (which are central to biologics and not easily reverse-engineered from the product), makes biologic patent thickets especially opaque compared with small-molecule drugs. (3) This opacity is one reason biosimilar litigation strategy diverges so sharply from generic small-molecule practice.