Hatch-Waxman Litigation: The Orange Book, Paragraph IV Certifications, and the 30-Month Stay

Document type: Article Practice area: Intellectual Property — Patents Jurisdiction: United States (federal) Last reviewed: 5 September 2026

Most patent litigation begins when somebody sells something. Hatch-Waxman litigation begins when somebody mails a letter.

That is not a metaphor. Under the statute, submitting an abbreviated new drug application with a particular kind of certification is itself an act of infringement — of a product that does not exist, cannot be sold, and may never be approved. Congress created this fiction deliberately, because the alternative was worse: without it, a generic manufacturer could not test the validity of a patent without launching at risk and facing catastrophic damages, and a brand manufacturer could not resolve the question until its market had already been taken.

The result is a litigation system with its own physics. There is no accused product. There are rarely damages. The remedy is a date. And a substantial part of the strategy happens at the Food and Drug Administration rather than in court.

For anyone who practices at the intersection of patents and regulated products, the framework repays careful study — not least because Congress has used the same architecture, with variations, for biologics.

The bargain

The Drug Price Competition and Patent Term Restoration Act of 1984 — universally called Hatch-Waxman — resolved a standoff.

The generic industry's problem. Before 1984, a company wanting to market a generic version of an approved drug had to run its own clinical trials, at enormous expense, to prove what was already known: that the drug is safe and effective. That made generic entry uneconomic for most products.

The brand industry's problem. Patent terms ran from issuance while regulatory review consumed years of the term. A company might obtain a patent, spend eight years in development and review, and have seven years of effective exclusivity for a fifteen-year investment.

The trade. Generics got the abbreviated new drug application — approval based on demonstrating bioequivalence to an approved reference product rather than on new clinical trials. Brands got patent term restoration for time lost to regulatory review, plus regulatory exclusivities independent of patents, plus a mechanism to litigate patent disputes before generic launch.

And the litigation framework exists to make the trade work: generics can challenge patents early, brands can enforce them before losing the market, and the dispute resolves on a schedule.

The Orange Book

The Approved Drug Products with Therapeutic Equivalence Evaluations — the Orange Book — is FDA's public list of approved drugs, their therapeutic equivalence ratings, and, critically, the patents the brand has listed as covering each drug.

What must be listed. Under 21 U.S.C. § 355, an NDA holder must submit patent information for patents that claim the drug substance, the drug product (formulation and composition), or an approved method of use, and with respect to which a claim of infringement could reasonably be asserted.

What must not be listed. Process patents, patents claiming metabolites, patents claiming intermediates, and — a point of longstanding contention — patents that do not claim the drug or an approved use at all. Device patents listed against drug-device combination products have drawn regulatory and enforcement attention.

Why listing matters so much. Listing is what triggers the certification requirement, which is what triggers the notice letter, which is what triggers the artificial act of infringement and the 30-month stay. A patent not listed in the Orange Book can still be enforced — but not through this framework, and without the stay.

Use codes. For method-of-use patents, the brand submits a use code describing the approved use the patent claims. The use code is what determines whether a generic can carve that use out of its label. Use codes that are broader than the patent actually claims can block a carve-out entirely, which is the problem Caraco Pharmaceutical Laboratories, Ltd. v. Novo Nordisk A/S, 566 U.S. 399 (2012) addressed — holding that a generic may bring a counterclaim to require the brand to correct a use code that inaccurately describes the patent's scope.

Delisting and the counterclaim. Section 355(j)(5)(C)(ii) permits an ANDA applicant sued for infringement to counterclaim seeking an order requiring the NDA holder to correct or delete patent information. It is a counterclaim, not an independent action, and it does not support damages.

The four certifications

An ANDA applicant must, for each patent listed against the reference drug, make one of four certifications under 21 U.S.C. § 355(j)(2)(A)(vii):

Paragraph I — no patent information has been filed. The application can be approved immediately as to patents.

Paragraph II — the patent has expired. Same.

Paragraph III — the patent has not expired, and the applicant seeks approval as of the date the patent expires. This is a concession: no litigation, no early entry, approval effective at expiry.

Paragraph IV — the patent is invalid, unenforceable, or will not be infringed by the drug for which approval is sought.

Paragraph IV is the one that starts a lawsuit.

Section viii statements. Distinct from a certification: where a listed patent claims a method of use that the applicant is not seeking approval for, the applicant may file a section viii statement and carve that indication out of its label. This is the "skinny label," and it is where much of the modern litigation happens. See below.

The artificial act of infringement

35 U.S.C. § 271(e)(2) provides that it is an act of infringement to submit an ANDA (or a paper NDA under § 505(b)(2)) for a drug claimed in a patent, if the purpose of the submission is to obtain approval to engage in commercial manufacture, use, or sale before the patent expires.

Why this is remarkable. No product has been made. Nothing has been sold. The "infringement" is a regulatory filing. Congress created subject matter jurisdiction for a case that would otherwise be unripe.

Eli Lilly & Co. v. Medtronic, Inc., 496 U.S. 661 (1990) confirmed the breadth of the companion provision, § 271(e)(1), holding that its safe harbor extends to medical devices as well as drugs — the Court reading the section as establishing a general framework rather than a drug-specific one.

What the infringement inquiry actually asks. Because there is no accused product on the market, the analysis compares the ANDA product as described in the application — the specification, the formulation, the proposed label — to the claims. The ANDA is the accused product. This makes the ANDA specification itself a central piece of evidence, and it makes what the applicant told FDA determinative in ways that surprise people new to the field.

The remedies are unusual. Under § 271(e)(4), the court may:

Order the effective date of approval to be a date not earlier than the date of patent expiration — the primary remedy, and it is a date, not money.
Grant injunctive relief against commercial manufacture, use, or sale.
Award damages only if there has been commercial manufacture, use, or sale — that is, only in an at-risk launch case.

The notice letter and the 30-month stay

The notice. An ANDA applicant making a Paragraph IV certification must give notice to the NDA holder and each patent owner, stating the factual and legal basis for the assertion that the patent is invalid or not infringed. This is the Paragraph IV notice letter, and it is the starting gun.

What it must contain. A detailed statement of the factual and legal basis for the applicant's opinion — in practice, a document running from tens to hundreds of pages, with claim charts, invalidity contentions, and an offer of confidential access to the ANDA.

The 45-day window. If the NDA holder or patent owner brings an infringement action within 45 days of receiving notice, FDA may not approve the ANDA for 30 months from the date of notice, or until a court decides the patents are invalid or not infringed, whichever is earlier.

The consequences of that clock:

The brand sues within 45 days, essentially always. Not suing forfeits the stay, and the stay is the primary benefit of the framework to the brand.
The complaint is filed on thin information. Forty-five days is not long enough to analyze a several-hundred-page notice letter, obtain the ANDA, and develop positions. Brands file protectively and refine later.
The 30-month period drives the schedule. Courts in ANDA-heavy districts manage cases toward decision before the stay expires, and the parties' scheduling proposals are framed by it.
The stay can be shortened or extended by the court if either party fails to cooperate reasonably in expediting the action.

Multiple stays. A brand that lists a new patent after the ANDA is filed cannot obtain a second 30-month stay on that patent — the statute permits only one stay per ANDA. This limits a practice that predated the 2003 amendments.

No preliminary injunction is needed. The stay does the work an injunction would do, without the brand having to show likelihood of success or irreparable harm. This is a significant structural advantage and it explains why the framework is valuable to brands even when their patents are weak.

180-day exclusivity

The generic incentive is a period of market exclusivity for the first applicant to file a substantially complete ANDA with a Paragraph IV certification.

What it provides. 180 days during which FDA may not approve a subsequent ANDA for the same drug. In practice this means the first filer competes only with the brand — and sometimes with the brand's authorized generic — during a period when generic pricing has not yet collapsed. The economic value is substantial, frequently the majority of the product's lifetime generic profit.

Shared exclusivity. Multiple applicants filing on the same day share first-filer status, which is why ANDA filings cluster on the first day of eligibility.

Forfeiture. The 2003 amendments added forfeiture events, including:

Failure to market within specified periods keyed to approval and to court decisions.
Withdrawal of the application.
Amendment of the certification.
Failure to obtain tentative approval within 30 months of filing.
Agreement with another applicant or the brand that violates the antitrust laws, as determined by a court or the FTC.
Expiration of all listed patents at issue.

Why forfeiture matters strategically. A first filer sitting on exclusivity blocks every subsequent filer — the "parking" problem the forfeiture provisions were designed to address. Subsequent filers watch the first filer's forfeiture triggers closely, and litigation sometimes targets them.

Authorized generics. A brand may launch or license its own authorized generic during the first filer's exclusivity, which substantially reduces the exclusivity's value. This is lawful and it is a standard brand response.

The safe harbor

35 U.S.C. § 271(e)(1) provides that it is not an act of infringement to make, use, offer to sell, or sell a patented invention solely for uses reasonably related to the development and submission of information under a federal law regulating the manufacture, use, or sale of drugs.

What it protects. All the pre-approval work: bioequivalence studies, stability testing, manufacture of validation batches, and the analytical work required to support the application.

How broad is "reasonably related"? Merck KGaA v. Integra Lifesciences I, Ltd., 545 U.S. 193 (2005)** read it broadly. The Court held that the safe harbor covers preclinical research on patented compounds where there is a reasonable basis to believe the compound tested could be the subject of an FDA submission and the experiments will produce the types of information relevant to an IND or NDA. It does not matter that the particular experiments are not ultimately submitted, or that the compound is not ultimately the subject of an application.

What it does not protect. Commercial manufacture and sale. Stockpiling commercial inventory before approval has been litigated, with courts examining whether the activity was reasonably related to a regulatory submission or was commercial preparation.

And note Eli Lilly v. Medtronic extends the safe harbor beyond drugs to medical devices and other FDA-regulated products, which matters far beyond the pharmaceutical context.

The skinny label and induced infringement

This is where the modern litigation risk concentrates, and it is worth understanding precisely.

The mechanism. A brand's drug is approved for indications A and B. A method-of-use patent covers indication B. A generic applicant files a section viii statement rather than a Paragraph IV certification as to that patent, and carves indication B out of its proposed label. It seeks approval only for indication A.

Why this exists. Congress wanted generic competition for unpatented uses even where a later-discovered use remains patented. The carve-out is the statutory mechanism.

The problem. Physicians prescribe by drug, not by indication. Pharmacists substitute by product. A generic approved for indication A is, in practice, dispensed for indication B — and the brand argues the generic induced that infringement.

What courts examine:

The label itself. Does anything remaining in the carved-down label point to the patented use — dosing information relevant only to the carved-out indication, a clinical study describing it, a pharmacology section that discusses it?
Marketing and promotional materials. Statements about therapeutic equivalence, or about the generic being the "same as" the brand.
Press releases and investor communications describing the product by reference to the full brand indication set.
Communications with formularies and payers.
What the generic knew about how the product would be used.

Where the doctrine sits. The Federal Circuit's decisions in this area — most prominently the GSK v. Teva litigation over carvedilol — have gone in more than one direction, with panel decisions, a rehearing, and dissents. The practical state of the law is that a skinny label can support induced infringement liability where the label or the generic's own communications encourage the patented use, and that a carve-out is not a categorical shield.

The practical instructions:

For generics: carve out completely, scrub every downstream document — labels, catalogs, websites, press releases, sales training, payer submissions — of any reference to the patented indication, and have counsel review the marketing plan before launch. Document the review.

For brands: collect the generic's public statements early and comprehensively. The case is built from the defendant's own materials, and they are usually public.

Reverse payment settlements

A Hatch-Waxman case settles in a way no other patent case does: the patentee pays the accused infringer, and the accused infringer agrees to stay out of the market until a date.

Why the shape is unusual. In ordinary patent litigation, the alleged infringer pays. Here, the brand has an enormously valuable monopoly, the generic has a lottery ticket, and both do better by splitting the monopoly rent than by litigating. The loser is the consumer who would have paid generic prices.

FTC v. Actavis, Inc., 570 U.S. 136 (2013) resolved the antitrust standard. The Court rejected both extremes: reverse payment settlements are not immune from antitrust scrutiny merely because they fall within the nominal scope of the patent, and they are not presumptively unlawful. They are subject to rule of reason analysis.

What the Court said matters:

A large and unjustified reverse payment can bring with it the risk of significant anticompetitive effects, and its size is a surrogate for the patentee's own assessment of its patent's weakness.
The antitrust analysis does not require litigating the patent's validity.
Justifications matter: payments for avoided litigation costs, for services with independent value, or for other legitimate consideration may explain the payment.

What has followed. Extensive private and FTC litigation over what counts as a payment (cash, but also side deals, supply agreements, and no-authorized-generic commitments), what counts as "large," and how to assess justification. Settlements are reportable to the FTC and DOJ under the Medicare Modernization Act, and the agencies review them.

The practical instruction. Any Hatch-Waxman settlement involving consideration flowing from brand to generic needs antitrust counsel involved in structuring it, not reviewing it afterward — and the file should document the litigation-cost and independent-value justifications contemporaneously.

How the case actually runs

ANDA litigation has a distinctive rhythm that surprises litigators from other patent practices.

No jury. These are bench trials. There are no damages in the ordinary case, so there is no Seventh Amendment right, and both sides typically prefer a judge for the technical issues.

No accused product. The ANDA is the accused product. Discovery centers on the application: the formulation, the manufacturing process, the stability data, the dissolution profiles, the proposed label. The ANDA itself is produced early, usually under an offer of confidential access made with the notice letter.

Venue is concentrated. Delaware and New Jersey handle a large share, for reasons of incorporation and industry geography. Judges in those districts have deep familiarity with the framework, published scheduling practices, and views on the recurring issues.

Multiple defendants, coordinated. A brand facing several ANDA filers typically sues all of them, and the cases are consolidated or coordinated for pretrial purposes. Defendants share invalidity work and split costs; each maintains its own non-infringement position because the ANDAs differ.

The schedule runs to the stay. A typical case: complaint at month 0, scheduling order by month 3, fact discovery closing around month 12–15, claim construction around month 12–18, expert discovery through month 20, summary judgment by month 22, and trial around month 24–28 — with a decision before the 30-month stay expires. Courts try to meet it; not all cases do.

The issues that recur:

Formulation patents: whether the ANDA formulation falls within claimed ranges; whether an excipient functions as claimed; the doctrine of equivalents where the ANDA is just outside a range.
Polymorph and crystalline form patents: what form the ANDA product actually contains, which turns on analytical testing of the applicant's own batches.
Method-of-use patents: whether the label instructs the claimed method; induced infringement; the skinny-label issues above.
Obviousness: whether the claimed formulation, salt, polymorph, or dosing regimen was obvious over the prior art, frequently with an obvious-to-try framing and unexpected-results rebuttal.
Written description and enablement for genus and range claims.
Obviousness-type double patenting, which does substantial work in this field where families are large and terminal disclaimers are common.

Secondary considerations carry real weight here. Commercial success, long-felt need, failure of others, and unexpected results are developed seriously in ANDA cases, and the brand's own launch and sales data are central evidence.

Regulatory exclusivities, which are not patents

Patent litigation is only half the exclusivity picture, and clients conflate the two constantly.

New chemical entity exclusivity — five years from approval of a drug containing no previously approved active moiety. No ANDA may be submitted during that period, except that an ANDA with a Paragraph IV certification may be submitted after four years.

New clinical investigation exclusivity — three years for a supplement or application containing new clinical investigations essential to approval. It blocks approval of an ANDA for that specific change, not the drug generally.

Orphan drug exclusivity — seven years for a designated drug for a rare disease, blocking approval of the same drug for the same use.

Pediatric exclusivity — six months added to existing patent and exclusivity periods in exchange for requested pediatric studies. Because it attaches to everything, it is often the most valuable exclusivity in the portfolio.

Why this matters in litigation. The relevant date is the latest of patent expiry (as extended) and applicable exclusivities. A brand that loses every patent may still have years of exclusivity; a brand that wins may find its patents expire before an exclusivity does. The settlement date is negotiated against the whole picture, not against a single patent.

Patent term extension under 35 U.S.C. § 156, restoring time lost to regulatory review, applies to one patent per product and is calculated by formula. It is frequently the single most valuable regulatory decision in a product's life, and it has its own litigation.

A worked timeline

Corvallis Therapeutics markets Dantrelex, approved for two indications: hypertension (indication A) and a rare cardiomyopathy (indication B). Three patents are listed: a compound patent expiring in 2031, a formulation patent expiring in 2034, and a method-of-use patent covering indication B expiring in 2036.

Month 0. Wexley Generics files an ANDA on the first day of eligibility, with:

A Paragraph IV certification to the formulation patent — Wexley contends its formulation is outside the claims and that the claims are obvious.
A Paragraph III certification to the compound patent — Wexley concedes it and seeks approval at expiry in 2031.
A section viii statement as to the method-of-use patent, carving indication B out of its label.

Month 0 + 20 days. Wexley sends its notice letter: 214 pages, claim charts, invalidity contentions, and an offer of confidential access to the ANDA.

Month 1. Corvallis obtains the ANDA under the confidential access agreement, and its formulation scientists begin analyzing the described product.

Month 1.5 (day 44). Corvallis files suit in the District of Delaware on the formulation patent. The 30-month stay attaches, running from the date of notice.

Corvallis also asserts the method-of-use patent on an induced infringement theory notwithstanding the carve-out, pointing to two things: Wexley's proposed label retains a dosing table that Corvallis contends is relevant only to indication B, and a Wexley investor presentation describes the product as "a generic version of Dantrelex."

Months 3–15. Discovery. The central factual disputes: what excipient ratio Wexley's product actually uses and whether it falls within the claimed range; what Wexley's batch records show; and what Wexley's marketing materials say.

Month 14. Claim construction. The dispositive term is a range limitation and whether it is measured before or after a processing step.

Month 18. Wexley moves for summary judgment of non-infringement on the formulation patent under the court's construction. Denied — a factual dispute about the measurement.

Month 25. Four-day bench trial.

Month 29. Decision. The formulation patent is not infringed — Wexley's product falls outside the claimed range as properly measured. The court does not reach validity. On the method-of-use patent, the court finds no induced infringement: the retained dosing table also applies to indication A, and the single investor statement is insufficient without evidence that Wexley encouraged the carved-out use.

Month 29 + 1 day. The 30-month stay is moot; FDA may approve.

Month 31. Wexley receives final approval and launches. As the first filer, it has 180 days of exclusivity. Corvallis launches an authorized generic on day one, which cuts Wexley's exclusivity value roughly in half.

Month 33. Corvallis appeals. Wexley remains on the market at risk of a damages award if the Federal Circuit reverses — the one circumstance in which § 271(e)(4) permits damages.

Month 45. Affirmed.

What each side did well. Wexley designed around the range and documented the design-around contemporaneously; it also scrubbed its marketing materials, and the single investor statement was the only slip. Corvallis sued within 45 days and preserved the stay, which bought it thirty months of exclusivity it would not otherwise have had — worth more than the litigation cost by an order of magnitude, even losing.

And note the asymmetry that defines this practice. Corvallis lost and still gained thirty months. That is the framework working as designed, and it is why brands assert patents here that they would not assert anywhere else.

The PTAB track

An ANDA filer can also challenge the listed patents at the Patent Trial and Appeal Board, and many do.

Why file an IPR alongside the district court case:

A different claim construction standard and a different burden — preponderance at the Board versus clear and convincing in court.
Speed. A final written decision is due within twelve months of institution, which can beat the district court.
A second bite on validity, with technical judges.

Why not:

Estoppel. 35 U.S.C. § 315(e) estops a petitioner who reaches a final written decision from asserting in district court any ground it raised or reasonably could have raised. This can eliminate the entire prior-art defense.
Discretionary denial. The Board may decline institution where parallel district court litigation is advanced, particularly where trial is scheduled before the Board's deadline. ANDA cases in fast districts are frequently in this posture.
It does not stop the clock. An IPR does not extend or shorten the 30-month stay, and FDA approval is not conditioned on it.
Eligibility is unavailable. Section 311 limits IPR to §§ 102 and 103 grounds on patents and printed publications.

The practical calculus. A generic with strong printed-publication prior art and a district court trial date beyond the Board's schedule may prefer the Board. One with a strong non-infringement position — which is unavailable at the Board — should think carefully, because estoppel can cost the invalidity defense in the forum where non-infringement will be decided.

Where the framework is under pressure

Four areas draw regulatory, legislative, and enforcement attention, and clients ask about all of them.

Orange Book listing practices. Listing patents that do not claim the drug or an approved method of use — device components of combination products has been the recent flashpoint — has drawn FTC warning letters and delisting demands, on the theory that improper listings extend the 30-month stay mechanism beyond its intended reach. The Caraco counterclaim and the delisting provisions of § 355 are the private mechanisms.

Patent thickets. Large families of formulation, device, and method patents around a single product can require an ANDA filer to certify to dozens of patents, each a potential litigation. Whether this is legitimate incremental innovation or exclusivity extension is genuinely contested, and the answer differs product by product.

Product hopping. Reformulating a product and shifting the market to the new version before generic entry on the old one has been the subject of antitrust litigation, with courts examining whether the reformulation had independent value and whether the transition coerced the market.

Settlement structures. Post-Actavis, agencies and private plaintiffs scrutinize non-cash consideration — supply agreements, licenses to other products, no-authorized-generic commitments — as potential reverse payments. The reporting requirement means every settlement is reviewed.

What this means practically. A brand building a lifecycle strategy and a generic evaluating a filing both need antitrust input at the strategy stage. The patent questions and the competition questions are not separable in this field.

Frequently asked questions

Why would a brand sue on a patent it expects to lose? Because filing within 45 days produces a 30-month stay of FDA approval regardless of the patent's strength. For a product with meaningful revenue, thirty months of exclusivity is worth far more than the litigation costs, and no showing of likelihood of success is required. This is a design feature of the statute, not an abuse of it — though it is the feature critics point to.

Can a generic launch before the case is decided? Yes, if FDA has approved the ANDA — which happens when the 30-month stay expires or a court holds the patents invalid or not infringed. Launching while an appeal is pending is an at-risk launch, and it is the one circumstance in which § 271(e)(4) permits damages, which can be enormous.

Does the 30-month stay stop the litigation? No. It stops FDA approval. The litigation proceeds, and courts try to decide before the stay expires.

What happens if the brand does not sue within 45 days? No stay. FDA may approve the ANDA when it is otherwise ready. The brand can still sue — the artificial act of infringement has occurred — but without the automatic delay, and it would need a preliminary injunction to stop a launch.

Can a brand get a second 30-month stay by listing a new patent? No. The statute permits one stay per ANDA. A newly listed patent requires a new certification from the applicant, and the brand can sue on it, but without a second stay.

Is a section viii carve-out safe? It is lawful and it is not a shield. Induced infringement liability turns on what remains in the label and what the generic says about the product. Carve out completely and scrub every downstream document.

Who is the first filer if several file on the same day? They share exclusivity. This is why filings cluster on the first eligible day.

Does an IPR help a generic? Sometimes. It offers a lower burden and a faster schedule, but it carries § 315(e) estoppel, may be discretionarily denied where the district court trial is earlier, does not affect the 30-month stay, and cannot reach non-infringement or eligibility.

How do these cases usually end? Settlement, on an agreed entry date, in the large majority of cases. The date is the deal, and the settlement is an antitrust event reportable to the agencies.

What is the single most important document? The ANDA. It is the accused product, and the applicant's own description of its formulation and process controls the infringement analysis.

The short version

The framework makes a paper filing an act of infringement, so that patent disputes resolve before generic launch rather than after.

Orange Book listing is the trigger. A patent not listed does not produce a certification, a notice letter, or a stay.

A Paragraph IV certification starts a lawsuit, and a suit filed within 45 days produces a 30-month stay of FDA approval — better than a preliminary injunction and available without showing likelihood of success.

The ANDA is the accused product. Discovery centers on the application, and what the applicant told FDA controls.

The remedy is a date, not money, unless there has been an at-risk launch.

180-day exclusivity drives generic behavior, and the forfeiture provisions and authorized generics shape its actual value.

The safe harbor of § 271(e)(1) is broad after Merck v. Integra and extends beyond drugs after Eli Lilly v. Medtronic.

Skinny labels are lawful and not a shield. Induced infringement liability turns on the label and the generic's own communications.

Settlements are antitrust events after FTC v. Actavis, analyzed under the rule of reason and reportable to the agencies.

And the structural point that explains the practice: a brand that loses on the merits after thirty months has still gained thirty months. The litigation is about a date, and both sides negotiate accordingly.

Section 505(b)(2): the third path

Between the full NDA and the ANDA sits a route that practitioners outside the field rarely know about and that generates a meaningful share of the litigation.

What it is. A § 505(b)(2) application is a new drug application that relies, in part, on data the applicant did not generate and does not own — typically FDA's findings of safety and effectiveness for a previously approved drug, or published literature.

When it is used. For a modification of an approved product: a new dosage form, a new strength, a new route of administration, a new indication, a different salt or ester, or a combination of approved actives. The applicant runs bridging studies rather than a full development program.

Why it matters here. A 505(b)(2) applicant must make the same certifications to Orange Book patents listed against the referenced drug, and a Paragraph IV certification triggers the same notice, the same artificial act of infringement under § 271(e)(2), and the same 30-month stay. The litigation looks the same.

The differences that matter:

There is no 180-day exclusivity for a 505(b)(2) applicant. The generic incentive structure does not apply.
The product is not a generic. It is not therapeutically equivalent and is not automatically substitutable, so it competes on the merits rather than at the pharmacy counter.
The applicant may obtain its own exclusivities — three-year new clinical investigation exclusivity for the change it studied, and its own patents on the modification.
The applicant is often a brand company, which makes the litigation brand-versus-brand and changes the settlement dynamics considerably.

The recurring dispute. Whether the 505(b)(2) product infringes patents covering the reference product, and whether the applicant's own modification is patentable over the reference product — the same prior art, run in both directions.

A practical note. Companies developing a modification of an approved drug frequently do not realize they are entering the Hatch-Waxman framework until counsel explains the certification requirement. Raise it at the development stage: the choice of reference product, the indications sought, and the formulation design all affect which patents must be certified to and whether a carve-out is available.