Introduction

FDA approval law rests on a statutory foundation short enough to quote in a paragraph. Section 505 of the Federal Food, Drug, and Cosmetic Act, codified at 21 U.S.C. 355, forbids the marketing of a new drug unless an approved application is effective, and section 351 of the Public Health Service Act, codified at 42 U.S.C. 262, licenses biological products on a showing of safety, purity, and potency. Nothing in either provision describes a clinical trial phase, a review clock, an advisory vote, or a fee. The operating system that sponsors, patients, and courts actually encounter is implementation: regulations, guidance, fee legislation, review organization, and post-marketing enforcement built around those two short commands across nine decades.

This article traces that operating system the way a practitioner meets it. A reader who completes it can follow a drug from the investigational application through the phases to approval and into post-marketing obligations, can explain why a substantial share of the review budget arrives as fees paid by the industry under review and why Congress renews that arrangement every five years, can describe the expedited pathways and the withdrawal problem that accelerated approval creates, and can state the preemption asymmetry that decides whether an injured patient can sue at all. The governing theme is the gap between statutory text and delivered outcome. The most consequential liability rule in the field emerged from the interaction of two statutes, neither of which addressed liability, and the most consequential funding rule lives in commitment letters rather than in the approval standard itself.

The Statutory Frame of FDA Approval Law and the Agency That Runs It

The core puzzle of FDA approval law lies in the distance between a short statutory command and the administrative system that gives that command practical force. Section 505 of the Federal Food, Drug, and Cosmetic Act, codified at 21 U.S.C. 355, states the basic rule for a new drug in compact terms: a person may not introduce or deliver for introduction into interstate commerce a new drug unless an approval of an application filed for that drug is effective. Section 351 of the Public Health Service Act, codified at 42 U.S.C. 262, states a parallel rule for a biological product through a license rather than an approval. Read in isolation, the two provisions suggest a gate, a gatekeeper, and a standard for opening the gate. Read against the delivered system, the provisions support an apparatus of investigational exemptions, staged human testing, manufacturing review, statistical assessment, labeling negotiation, facility inspection, expert committees, and postsubmission correspondence. The statute supplies the legal skeleton. Regulation, guidance, fee legislation, scientific custom, and agency organization supply much of the muscle.

That distance matters because Congress enacted no complete examination system in one text. The Federal Food, Drug, and Cosmetic Act identifies the prohibited act, the required application, the principal grounds for refusal, and the statutory meaning of substantial evidence. The Public Health Service Act identifies the license, names safety, purity, and potency as the qualities to be demonstrated, and directs attention to the manufacturing facility. Neither provision explains how a sponsor assembles proof, how agency specialists divide an application, or how a review reaches a decision on a predictable schedule. Those questions belong to implementation.

The statutory text also speaks through the Secretary of Health and Human Services. Section 505 directs findings by the Secretary and requires an order approving an application when the statutory grounds for refusal do not apply. The Food and Drug Administration, an agency within the Department of Health and Human Services, performs the work through delegated authority and specialized centers. The formal chain therefore runs from Congress to the Secretary, from the Secretary to the Food and Drug Administration, and from the agency to review divisions and discipline teams. An approval bears institutional authority rather than the personal judgment of a single reviewer, even though named officials sign decisions and individual reviewers write the assessments on which those decisions rest.

The Formal Identity of an Approval and a License

A new drug approval and a biologics license answer related but distinct legal questions. The new drug application asks whether a drug may enter interstate commerce for the conditions of use prescribed, recommended, or suggested in its proposed labeling. The inquiry is use specific. Evidence supporting one dose, one route of administration, one population, or one indication does not automatically support another. The application contains reports of investigations, a statement of the drug components and composition, a description of manufacturing methods and controls, samples and labeling material as required, and other information prescribed by the governing provisions. Approval makes the application effective for the product and labeling reviewed, rather than conferring a general endorsement of a chemical entity for every purpose a sponsor or prescriber might imagine.

The biologics license application asks whether a biological product is safe, pure, and potent and whether the facility connected with the product meets standards designed to assure that the product continues to be safe, pure, and potent. The facility element has unusual weight for biologics because many such products derive from living sources, depend on tightly controlled production conditions, and resist complete characterization through the finished package alone. Section 262(a) therefore joins product demonstration to manufacturing setting and to consent for inspection. A license is not a finding that an abstract substance performed well in a trial detached from production. The license rests on a product made by an identified process in a facility subject to agency inspection, with labeling and package marking rules forming part of the statutory scheme.

The two routes overlap without collapsing into one another. A biological product can also fall within the drug provisions of the Federal Food, Drug, and Cosmetic Act, and licensure interacts with prohibitions on adulteration and misbranding. Assignment inside the Food and Drug Administration does not follow the license label in a mechanical way either, since the Center for Drug Evaluation and Research handles many therapeutic proteins, including monoclonal antibodies, while the Center for Biologics Evaluation and Research handles vaccines, blood products, allergenics, tissues, and cellular and gene therapies. The legal category, the product characteristics, and the agency assignment each contribute to the path an application follows. Approval and license are not interchangeable names for a single act. Each term records a statutory source and a set of expectations about manufacturing and proof.

Safety entered the federal premarket system before effectiveness. The Federal Food, Drug, and Cosmetic Act of 1938 required a showing of safety for a new drug, built around adequate tests by methods reasonably applicable to the proposed conditions of use. The Act did not impose the later general requirement that effectiveness be proved through substantial evidence before marketing. That effectiveness limb entered section 505 through the amendments that added the effectiveness requirement, which left safety in place and added a distinct evidentiary gate for the effect a drug purports or is represented to have. A compressed statement of the result is enough for the operating frame: after 1962, section 355(d) required both a safety basis and substantial evidence of effectiveness, and a failure on either ground supported refusal.

What does substantial evidence mean under FDA approval law?

Substantial evidence means effectiveness evidence drawn from adequate and well-controlled investigations, including clinical investigations, conducted by qualified experts, from which those experts could fairly and responsibly conclude that a drug will produce the effect claimed in its proposed labeling for the stated conditions of use.

The statutory definition performs several tasks at once. The phrase adequate and well-controlled investigations directs attention to design rather than volume alone, since a large accumulation of testimonials or uncontrolled observations does not become substantial merely through repetition. The reference to experts qualified by scientific training and experience ties the conclusion to disciplined judgment. The phrase fairly and responsibly places a reasonableness condition on the inference from data to claimed effect, and the closing reference to conditions of use in labeling confines the conclusion to the use under review. Regulation at 21 C.F.R. 314.126 later elaborated the characteristics of an adequate and well-controlled study, including a clear statement of objectives, a design permitting a valid comparison with a control, measures intended to minimize bias, and reliable methods of assessing response.

Substantial evidence is therefore neither a synonym for certainty nor a fixed count of trials written into section 355(d). The statute defines an evidentiary quality and assigns the Secretary the task of deciding whether an application meets that quality on the record submitted and on other information before the agency. Amendments and agency practice have allowed reliance on one adequate and well-controlled clinical investigation together with confirmatory evidence in appropriate circumstances, a development that changes the composition of a record without abandoning the governing standard. The distinction is subtle and consequential. Flexibility concerns the kinds and combination of proof capable of satisfying the definition. Flexibility does not convert a plausible mechanism, a favorable case series, or a sponsor assertion into substantial evidence by relabeling the material.

Safety under section 355(d) follows a different formulation. The refusal grounds ask whether the investigations include adequate tests by all methods reasonably applicable to show whether the drug is safe for use under the proposed labeling conditions, whether the results show safety or fail to show safety, and whether the information remains insufficient for a safety determination. Safety in this setting has never meant the absence of all risk, because the statutory question attaches risk to a stated use, dose, population, and labeling framework. A drug with serious hazards may present an acceptable safety profile for a life-threatening condition with adequate warnings and controls, while a lesser hazard may defeat a proposed use offering marginal benefit for a minor condition. The text separates the safety clauses from the substantial evidence clause, preserving a distinction between proof that a product works and judgment about the conditions under which exposure is acceptable.

Two Centers and the Organization of Expertise

The Food and Drug Administration divides the human drug and biologics workload principally between the Center for Drug Evaluation and Research and the Center for Biologics Evaluation and Research. The Center for Drug Evaluation and Research, known by the initials CDER, carries responsibility for most prescription and nonprescription drugs, generic drugs, and many therapeutic biological products transferred into the drug center framework. The Center for Biologics Evaluation and Research, known by the initials CBER, carries responsibility for vaccines, blood and blood products, allergenic products, human tissues, and cellular and gene therapies, together with related products assigned to that center. Both centers sit inside one agency and operate under one Commissioner, yet each center developed laboratory traditions, review cultures, and product offices suited to different scientific problems.

The division of labor follows risk and method as much as legal label. A chemically synthesized small molecule ordinarily presents questions of molecular identity, impurities, dose response, metabolism, and reproducible manufacture at scale. A vaccine presents questions of antigen characterization, immune response, lot consistency, and prevention endpoints in populations that may include healthy children. A cellular therapy presents questions of donor material, cell processing, viability, chain of custody, and a manufacturing run that may be created for an individual patient. Placing vaccines and blood in one center and most conventional drugs in another allows reviewers and laboratory scientists to accumulate knowledge around recurring production methods and failure modes. Jurisdictional questions at the border remain possible, particularly for combination products and novel platforms, and the agency maintains mechanisms for assigning a lead center according to the primary mode of action and the expertise required.

Within CDER, the Office of New Drugs forms the principal home for review of investigational new drug programs and new drug applications for innovator products. Review divisions inside that office are organized by therapeutic area, so that applications cluster around cardiology, neurology, psychiatry, oncology, infectious disease, endocrinology, gastroenterology, and other fields of clinical practice. The arrangement places a proposed treatment near reviewers who follow the disease literature, understand accepted endpoints in that field, recognize the standard therapies against which benefit will be judged, and see patterns across competing development programs. A division handling products for a rare neurological disorder confronts natural history data, surrogate endpoint disputes, and small trial populations repeatedly, while a division handling anti-infective products confronts resistance patterns, microbiological endpoints, and trial designs shaped by the ethics of withholding effective therapy. Therapeutic organization converts experience into institutional memory.

Staffing a Single Application

A single marketing application is staffed as a multidisciplinary review rather than as a reading assignment for one official. A medical officer examines the clinical investigations, the selection of endpoints, the handling of missing data and protocol deviations, the pattern of benefit across subgroups, and the adverse events that define the safety profile. A pharmacologist or toxicologist evaluates animal pharmacology and toxicology, asks whether the nonclinical findings predict human hazards, and considers whether the clinical program addressed signals arising before human exposure. A chemist and other product quality reviewers examine composition, controls, stability, specifications, and the manufacturing process, including whether the commercial process will reproduce the material used in pivotal investigations. A statistician interrogates randomization, blinding, prespecification, multiplicity, interim analysis, missing outcome handling, and the sensitivity of conclusions to analytic choices. Clinical pharmacology reviewers address absorption, distribution, metabolism, excretion, dose selection, organ impairment, and interactions with other drugs. Depending on the product, microbiologists, epidemiologists, device specialists, labeling experts, and facility inspectors add further assessments.

The disciplines meet through a review team and a regulatory project manager who coordinates meetings, correspondence, deadlines, and the assembly of the administrative record. Each discipline prepares a written review that identifies the material examined, states conclusions, and explains disagreements or unresolved questions. Supervisors and cross-discipline leaders test those conclusions against the whole record, and the signatory official for the responsible division or office makes or recommends the action under the applicable delegation. The architecture serves a checking function. A statistician may find that an apparently favorable endpoint loses force after correction for multiple comparisons. A chemist may find that a scale up change weakens the bridge between trial material and commercial material. A medical officer may accept a safety signal as manageable only after labeling and dose restrictions reshape the proposed conditions of use. The final action letter compresses those separate inquiries into an agency decision, but the underlying record preserves the plural character of the judgment.

The Interface With Drug Scheduling

Approval law and drug scheduling law ask different questions about overlapping evidence. An approval decision asks whether a specific product, made in a specific way and labeled for specific conditions, has the safety basis and substantial evidence required for marketing. A scheduling decision under the Controlled Substances Act asks where a substance belongs within a control system built around abuse potential, dependence liability, public health risk, pharmacological effect, scientific knowledge, history and pattern of abuse, and accepted medical use. The Department of Health and Human Services supplies a scientific and medical evaluation and a scheduling recommendation in the administrative scheduling process, with the Food and Drug Administration performing the scientific work inside the department and consulting the National Institute on Drug Abuse. Recommendations on scientific and medical matters bind the Drug Enforcement Administration in that process, while the final scheduling rule remains an action of the Attorney General through delegated authority.

The junction appears most clearly in the scheduling statute’s treatment of accepted medical use, which makes medical acceptance one factor in placement rather than a complete answer to placement. An approved drug may remain in Schedule II, Schedule III, Schedule IV, or Schedule V when abuse potential and dependence risks justify control despite recognized therapeutic value, as the placement of many approved opioid, stimulant, sedative, and anticonvulsant products demonstrates in the scheduling lists. Conversely, a substance lacking an approved marketing application presents a different evidentiary posture on medical acceptance, and the scheduling inquiry draws on scientific literature, clinical experience, state level practice, expert evaluation, and the statutory factors rather than on approval status alone. FDA approval law therefore feeds scheduling without governing scheduling. A license or approval establishes that a defined product met a marketing standard for defined uses. Scheduling classifies a substance for control purposes across a broader field of abuse and diversion concerns.

Where the Statute Falls Silent

The silences in section 505 and section 351 are as instructive as the commands. Section 505 does not prescribe the familiar sequence of clinical trial phases. The phase vocabulary and the staged description of investigations arise principally from the investigational new drug regulations, notably the framework associated with 21 C.F.R. Part 312, together with protocol practice and guidance. An early phase investigation centered on safety, dosage, and pharmacokinetics answers a different question from a later controlled investigation centered on effectiveness and adverse reactions in the target population, yet the statute itself speaks in broader terms of investigations, tests, and substantial evidence. Congress set the destination for proof and left the route largely to regulation and scientific method. Later implementation sections can therefore treat phases as an administrative and scientific construction built under the statute, rather than as a timetable dictated by the statutory words.

Section 505 likewise contains no general review clock of the kind associated with application practice under the fee system. The predictable goal dates for agency action grew out of the Prescription Drug User Fee Act of 1992 and the series of reauthorizations that followed in 1997, 2002, 2007, 2012, and 2017. Congress used fee authority to fund additional review capacity and incorporated performance goals through findings and commitment letters rather than writing a universal deadline into the original approval standard. The Congressional Research Service account of the 1992 legislation describes goals under which the agency was to review and act on most standard and priority applications within stated periods, with later cycles refining the measures. A missed goal date and a statutory refusal ground occupy different legal ground. The clock disciplines the institution and shapes applicant planning, while section 355(d) supplies the findings required for approval or refusal on the merits.

Advisory votes occupy a similar position outside the core text. Committees of independent experts, organized under the Federal Advisory Committee Act framework formalized in 1972, advise on questions framed by the agency after public presentations and questioning. Section 505 does not make a favorable committee vote an element of substantial evidence and does not transfer the approval decision to a committee. The recommendation does not bind the Food and Drug Administration. Guidance documents occupy yet another layer. A guidance states agency thinking on endpoints, trial design, manufacturing expectations, or data presentation, and guidance under the Food and Drug Administration good guidance framework does not itself create legally binding duties in the manner of a statute or a regulation adopted through rulemaking. Applicants ignore guidance at practical peril because guidance reveals the questions reviewers will ask, yet the legal obligation remains anchored in the statute and binding regulations.

Rulemaking fills other gaps through public procedure under the Administrative Procedure Act. Regulations define application contents, elaborate adequate and well-controlled investigations, govern investigational exemptions, set biologics standards, and prescribe procedures for approval or refusal. Section 262(a) expressly directs the Secretary to establish requirements for approval, suspension, and revocation of biologics licenses by regulation, making delegation visible on the face of the license provision. Section 701 of the Federal Food, Drug, and Cosmetic Act supplies general rulemaking authority for efficient enforcement of that Act. The result is a hierarchy that later sections must keep in view: statute controls regulation, regulation controls guidance, and an individual approval applies the whole hierarchy to a single record.

The frame for the sections that follow is consequently a layered one. At the center stand two short propositions: a new drug needs an effective approval supported by the safety showing and substantial evidence demanded by 21 U.S.C. 355, and a biological product needs a license supported by a demonstration of safety, purity, and potency under 42 U.S.C. 262. Around those propositions stand the centers, offices, divisions, review disciplines, inspectors, and committees that turn propositions into decisions. Beyond the agency stand fee commitments that set expectations for timing, regulations that define acceptable investigations, and guidance that maps preferred methods for reaching the statutory destination. FDA approval law operates in the interaction among those layers, and any account confined to the statutory lines alone would mistake the doorway for the building.

From Investigational Exemption to Market Authorization

The pathway to lawful marketing runs through two distinct legal permissions. The first allows a sponsor to test an unapproved product in human beings under controlled conditions. The second allows the sponsor to sell the product to the public. Neither follows automatically from the other. A sponsor may complete years of testing and receive no marketing authorization, because testing authority answers whether an investigation may proceed, while approval answers whether the evidence satisfies the statutory standard for safety and effectiveness. Section 505(a) of the Federal Food, Drug, and Cosmetic Act, codified at 21 U.S.C. 355(a), prohibits introduction of a new drug into interstate commerce unless an approval is effective. The investigational provisions carve a conditional space inside that prohibition, in which shipment and administration are lawful for investigation alone and in which the Food and Drug Administration retains authority to stop the work on safety grounds.

The Preclinical Record That Precedes Human Testing

Before any human being receives an investigational drug, the sponsor must assemble a preclinical record, and the legal function of that record is narrower than a general showing that a compound looks promising. The record rests principally on pharmacology and toxicology work in laboratory animals and on in vitro studies, together with information about the composition, manufacture, and control of the product to be administered. Animal studies establish how the compound behaves in a living system, how it is absorbed and distributed and eliminated, and at what exposures injury appears in organs such as the liver, the kidney, and the heart. Toxicology studies of differing durations, including studies addressed to reproductive effects, to genetic injury, and, for products intended for prolonged use, to carcinogenic potential, supply the margin of safety reasoning on which a first human dose is selected.

The preclinical record also forces the sponsor to organize, before human exposure begins, the information investigators and the agency need in order to reason about risk: the rationale for the study, prior experience with the drug or with related compounds, known and suspected hazards, and the precautions the protocol builds around those hazards. That information later appears in condensed form in the investigator brochure, which allows an investigator who did not design the program to understand why the study is being run and what dangers require monitoring. The standard at this stage is not proof of effectiveness and not proof of absolute safety. The standard is a record adequate to justify exposing a defined group of human subjects, under a defined protocol, to a risk characterized as honestly as the available science permits.

The Investigational Application and the Power to Impose a Clinical Hold

The investigational new drug application, shortened to IND, is the instrument through which the exemption from the marketing prohibition is claimed and policed. Its statutory basis lies in section 505(i) of the act, codified at 21 U.S.C. 355(i), which authorizes the agency to exempt drugs intended solely for investigational use from the approval requirement, subject to conditions prescribed by regulation at 21 C.F.R. Part 312. The application, described in 21 C.F.R. 312.23, assembles the preclinical record, the proposed clinical protocol or protocols, information about the investigators, commitments to obtain informed consent and institutional review board oversight, and the investigator brochure. Submission grants nothing by itself. Under 21 C.F.R. 312.40, an investigation may ordinarily begin thirty days after the agency receives the IND. The agency uses that interval to review the submission for safety concerns, and if the agency takes no blocking action, the investigation becomes effective by operation of the regulation rather than by an affirmative grant.

The counterweight to default effectiveness is the clinical hold, governed by 21 C.F.R. 312.42. A clinical hold is an order directing the sponsor to delay a proposed investigation or to suspend an investigation in progress. When a proposed study is placed on hold, subjects may not receive the investigational drug. When an ongoing study is placed on hold, new subjects may not be recruited into treatment, and subjects already receiving the drug are ordinarily taken off that therapy unless the agency permits continuation in the interest of patient safety. The grounds for a hold in a Phase I study include exposure or prospective exposure of subjects to an unreasonable and significant risk of illness or injury, investigators who lack the scientific training and experience to conduct the study, an investigator brochure that is misleading or erroneous or materially incomplete, and an application that lacks sufficient information for the agency to assess the risks to subjects. For later phases, the regulation adds a ground addressed to defective design, reaching a study plainly deficient in design to meet its stated objectives. A hold ends only when the agency lifts it, ordinarily after the sponsor submits a complete response to the stated deficiencies.

The Early Human Studies and the Question of Tolerable Exposure

Phase I is the first period of human testing under an effective IND, and its legal function is to generate the safety and dosing information without which larger human exposure would be irresponsible. The regulation at 21 C.F.R. 312.21 describes Phase I as the period in which a drug is initially introduced into human beings, with emphasis on the metabolic and pharmacologic actions of the drug, the side effects associated with increasing doses, and, where the design permits, early evidence of effectiveness. Studies are closely monitored and may be conducted in patients or in normal volunteer subjects. Numbers are small by design, commonly involving a total in the range of twenty to one hundred subjects across the phase. Dose escalation proceeds in small steps, advancing only after review of the experience at the preceding level. The output of Phase I is not a claim that the drug works. The output is a foundation: a dosing range human beings tolerate, a description of how the body handles the compound, a catalog of adverse effects later protocols must monitor, and a basis for choosing doses for controlled testing.

Phase II follows when the early safety record supports controlled testing in patients who have the disease or condition under study. The regulation describes Phase II as comprising controlled clinical studies conducted to evaluate effectiveness for a particular indication and to determine the common short term side effects and risks in a larger group than Phase I exposed. Numbers grow from dozens to, in many programs, a few hundred subjects. The legal function of the phase is refinement with discipline. Dose ranging studies compare several doses against one another, and sometimes against a control, in order to identify the dose or doses that carry a favorable balance into large scale testing. Endpoints are tested for feasibility, populations are narrowed or broadened as the evidence instructs, and the sponsor learns whether the effect hoped for in patients appears under controlled conditions.

Large Controlled Investigations and the Meetings That Frame Them

Phase III is the expanded controlled testing that ordinarily supplies the core of the approval case. Under 21 C.F.R. 312.21, Phase III follows a preliminary indication in Phase II that the drug may be effective, and it is conducted in a larger patient population to gather additional information about safety and effectiveness, to evaluate the overall benefit and risk relationship, and to provide an adequate basis for physician labeling. The statutory demand behind the phase is the substantial evidence standard: approval requires substantial evidence of effectiveness, defined by reference to adequate and well controlled investigations conducted by qualified experts, on the basis of which qualified experts could fairly and responsibly conclude that the drug will have the effect claimed for it. The characteristics of an adequate and well controlled investigation, set out at 21 C.F.R. 314.126, include a clear statement of objectives, a design that permits a valid comparison with a control, subject selection methods that provide reasonable assurance that subjects have the condition under study, assignment procedures that minimize bias, measures to minimize bias in observation and analysis, well defined and reliable methods of assessing response, and an analysis adequate to assess the effects of the drug.

Because Phase III is expensive, lengthy, and decisive, the pathway surrounds it with formal meetings whose function is to reduce the risk that sponsor and agency are building toward different judgments of what the evidence must show. The end of Phase II meeting, provided for in 21 C.F.R. 312.47, holds the greatest practical weight. At that meeting the sponsor presents the Phase I and Phase II results and the proposed Phase III program, and the parties address whether the planned studies are adequate in design, size, and endpoint selection to support a marketing application, what the safety database should contain, and whether additional work on dose or population must precede large trials. Agreement at that stage is not a promise of approval, but a sponsor that ignores the recorded advice of the review division proceeds at considerable peril. The pre-application meeting, held before submission of the marketing application, turns from study design to presentation: the format and organization of the application, the analyses the agency expects, the labeling concept, and any unresolved issue that might support a refusal to file.

The Marketing Application for Drugs and Biologics

When the sponsor judges the evidence ready, the legal posture changes completely. The sponsor is no longer seeking room to investigate; the sponsor is asking the government to authorize sale, and the burden of persuasion rests entirely on the applicant. For a drug, the vehicle is the new drug application under section 505(b) of the act, codified at 21 U.S.C. 355(b). Section 355(b)(1) prescribes the contents: full reports of investigations made to show whether the drug is safe for use and effective in use; a full list of the articles used as components; a full statement of composition; a full description of the methods used in, and the facilities and controls used for, manufacture, processing, and packing; samples where the agency requires them; specimens of proposed labeling; and the patent information the statute specifies. The application regulations at 21 C.F.R. 314.50 organize that material into sections on chemistry, manufacturing, and controls, nonclinical pharmacology and toxicology, human pharmacokinetics and bioavailability, microbiology where relevant, clinical data, safety updates, statistics, and an integrated summary in which the applicant argues the benefit and risk case as a whole.

Biological products follow a parallel track under a different statute. A biologics license application proceeds under section 351(a) of the Public Health Service Act, codified at 42 U.S.C. 262(a), and the license issues on a demonstration that the product is safe, pure, and potent and that the facility in which the product is manufactured meets standards designed to assure that the product continues to be safe, pure, and potent. Potency in that formulation performs the work that effectiveness performs for drugs, expressing the capacity of the product to achieve its intended result. The biologics framework places distinctive weight on the manufacturing process, because a living production system can alter the product in ways chemical synthesis ordinarily does not, and reviewers therefore examine process validation, characterization, and facility control with particular care.

The Filing Decision and the Refusal to File

Receipt of a marketing application starts no substantive review until the agency decides that the application may be filed, and that threshold decision has its own regulation and deadline. Under 21 C.F.R. 314.101, the agency determines within sixty days after receipt whether a new drug application may be filed, and filing signifies a threshold determination that the application is sufficiently complete to permit substantive review. The date of filing is the sixtieth day after receipt. A parallel filing framework applies to biologics license applications under 21 C.F.R. 601.2. If the application is on its face incomplete, lacking information required by the statute and the application regulations, the agency may refuse to file it. A refuse to file decision is not a judgment on the merits. The agency states the reasons in writing, and the applicant may amend and resubmit, or, in the circumstances the regulation describes, may request an informal conference and then ask that the application be filed over protest, in which posture the agency reviews the application as filed and the deficiencies remain in the record.

How long does the federal review of a marketing application take?

For an original application, the performance goal is ten months from filing for a standard review and six months from filing for a priority review. Filing occurs sixty days after receipt, after the agency decides that the application is sufficiently complete for substantive review.

Those figures require careful explanation, because the review clock is a performance commitment rather than a statutory deadline with automatic consequences. The goals arise from the user fee program, under which applicants pay fees that support the review process and the agency commits, in commitment letters negotiated for each reauthorization period, to act on stated percentages of applications within stated intervals. For original new drug applications and biologics license applications, the framework assigns a goal of ten months from the filing date for a standard application and six months from the filing date for a priority application. Priority review designation rests on the agency judgment, made at filing, that the product, if approved, would represent a significant improvement in the safety or effectiveness of the treatment, diagnosis, or prevention of a serious condition. The sixty days between receipt and filing sit outside the stated goal interval, so elapsed time from submission to action ordinarily exceeds the headline figure by roughly two months. Missing the goal date does not approve the application by default and does not extinguish agency authority.

Inside that interval, review proceeds as a multidisciplinary audit rather than as a single reading. The review team ordinarily includes a medical officer who evaluates the clinical trials and the benefit and risk balance, a pharmacology and toxicology reviewer who revisits the animal record against the human experience, a chemistry reviewer who examines drug substance and drug product manufacture and control, a biopharmaceutics reviewer who addresses absorption and formulation performance, a statistician who tests whether the analyses support the conclusions drawn from them, and, as the product requires, microbiology, clinical pharmacology, and other specialists. Reviewers issue discipline level assessments, consult at internal meetings, and send information requests when the record contains gaps clarification might close. Late in the cycle, reviewers turn to labeling, negotiating the precise claims, warnings, and dosing instructions the evidence supports, and to inspection of the manufacturing facilities named in the application. An application can fail on manufacturing grounds alone, because the statute conditions approval on adequate methods, facilities, and controls as an independent showing.

Some applications are referred to an advisory committee, a body of outside experts convened under the Federal Advisory Committee Act, supplemented by statutory provisions that require referral or an explanation for its absence for certain applications. The committee hears presentations from the applicant and from the agency review team in a public meeting, questions both sides, and votes on questions the agency frames, which may address effectiveness, safety, and the ultimate benefit and risk judgment. The vote does not bind the agency.

The Approval Letter and the Complete Response Letter

Review ends in one of two letters, and the difference between them defines the applicant position after the clock runs out. An approval letter under 21 C.F.R. 314.105 states that the application is approved, identifies the labeling and any postmarketing requirements or commitments that attach, and authorizes marketing of the product as labeled from the date of the letter. The complete response letter, governed by 21 C.F.R. 314.110, issues when the agency has completed review and determines that the application cannot be approved in its present form. That letter describes all of the specific deficiencies the agency has identified, identifies amendments the agency has not yet reviewed where relevant, and, when possible, recommends actions the applicant might take to place the application in condition for approval. The letter replaced the older approvable and not approvable letters after the agency amended the regulation in a rulemaking completed in 2008.

A complete response letter is not a denial, and treating it as one misstates the obligations it creates. Under 21 C.F.R. 314.110(b), the applicant must take one of three actions: resubmit the application, addressing all deficiencies identified in the letter; withdraw the application, a step taken without prejudice to a later submission; or request an opportunity for a hearing on the question whether grounds exist for denying approval. The resubmission path is the one most applicants take, and it starts a new review cycle on receipt, with the agency classifying the resubmission by the scale of the material it contains and assigning a corresponding review goal. If the applicant takes none of the three actions within one year after issuance of the letter, the agency may consider the failure to be a request to withdraw the application, unless the applicant has obtained an extension of time to address the deficiencies. The agency follows a notice procedure that allows the applicant thirty days to explain why the application should not be withdrawn and to seek an extension, with reasonable extension requests granted. What the letter obliges, at minimum, is a decision, because the regulation converts silence into withdrawal.

The Abbreviated Route and the Middle Route

For a generic version of an approved drug, Congress provided an abbreviated new drug application under section 505(j) of the act, and the compressed logic of that route is that the applicant relies on the agency prior finding of safety and effectiveness for the listed drug it copies, substituting a showing that the generic product has the same active ingredient, dosage form, route of administration, and strength, together with evidence of bioequivalence and labeling that is the same in the respects the statute requires. The certification and notice machinery that governs patent rights within that route, and the litigation stay that a patent suit can trigger, belong to the design of the statute that created the pathway, and a full account of the abbreviated pathway and its certification machinery treats that machinery on its own terms. Within the pathway account, the point is structural. The abbreviated applicant conducts no independent program of investigation to prove safety and effectiveness from first principles, the filing screen operates as a determination whether the submission is substantially complete and may be received, and review ends in approval, tentative approval where a legal barrier postpones final effectiveness, or a complete response letter carrying the same three options of resubmission, withdrawal, or a hearing request.

Between the full application and the abbreviated one sits the route of section 505(b)(2), the middle path of the system. An application of that type is a new drug application that relies, for some part of the safety and effectiveness case, on investigations the applicant did not conduct or sponsor and for which the applicant has no right of reference, most often published literature or the agency prior findings for an approved drug. The route suits products that differ from an approved drug in a way that defeats the sameness requirements of the abbreviated pathway, such as a changed dosage form, a new strength, a modified formulation, a different route of administration, or a new use, while sharing enough with the approved product that repeating the entire evidentiary program would waste resources and expose subjects to investigation without scientific need. The applicant must supply the bridge: scientific evidence, frequently including bioavailability or bioequivalence comparisons and any additional clinical work the differences require, connecting the relied upon findings to the product actually proposed. Because the application is a full new drug application in form, it passes through the same filing decision, the same review clock and goal framework, the same multidisciplinary review, and the same pair of possible action letters as an application built entirely on original studies.

The pathway and obligation table

The stages described in this section and the sections that follow reduce to a single sequence, in which each legal instrument creates the obligation the next stage enforces.

Stage	Legal instrument	Agency action	Timeline	Obligation created
Preclinical investigation	Sponsor record under the investigational framework	None required before submission	Before any human exposure	A record adequate to justify first human dosing
Investigational application	21 U.S.C. 355(i) and 21 C.F.R. Part 312	Safety review, with clinical hold available	Effective 30 days after receipt unless a hold issues	Informed consent, institutional review, and protocol compliance
Early phase testing	21 C.F.R. 312.21	Monitoring through reports and hold authority	Staged dose escalation	Safety and dosing data adequate for controlled trials
Controlled investigations	21 C.F.R. 312.21 and 314.126	End of phase meetings and design advice	The longest stage of development	Adequate and well controlled evidence of effectiveness
Marketing application	21 U.S.C. 355(b) or 42 U.S.C. 262(a)	Filing decision	Decision within 60 days of receipt	A complete record supporting approval or refusal
Agency review	User fee performance goals	Multidisciplinary review and possible advisory referral	Goal of 10 months standard and 6 months priority from filing	Labeling negotiation and facility inspection
Approval	Approval letter under 21 C.F.R. 314.105	Authorization to market as labeled	Effective on the date of the letter	Post-marketing duties attach
Post-marketing period	21 U.S.C. 355(o)(3) and 355-1; 21 C.F.R. 314.80	Signal detection, required studies, and risk strategies	Continuing for the life of the product	Adverse event reporting and study completion
Accelerated approval confirmation	21 U.S.C. 356(c)	Progress reports every 180 days after the 2022 reform	Study milestones set at approval	Verification of predicted clinical benefit
Withdrawal	21 U.S.C. 355(e) and 21 C.F.R. 314.530	Notice, sponsor participation, and decision	Expedited procedures codified in 2022	Removal of the approval or the indication

The User Fee Funding Structure

Approval law in the United States runs on money, and for prescription drugs the money arrives through two channels that Congress braided together in 1992. One channel is the ordinary appropriation, voted by Congress and drawn from the Treasury. The other is a set of fees paid by the companies whose applications the Food and Drug Administration reviews. The fee channel transformed the economics of review. It gave the agency a workforce sized to the workload, gave sponsors a predictable clock, and gave Congress a recurring negotiation in which funding, performance, and drug policy are traded in a single package. The structure also created the central controversy of modern approval law, because the regulator of an industry came to depend on payments from that industry for a large and growing share of the cost of regulation.

Why Congress Turned to Fees Paid by Sponsors

By the early 1990s the review system was failing on the dimension patients noticed most, which was time. New drug applications waited in queues that the agency lacked the staff to clear, and average review times in the late 1980s ran well beyond two years for many products, with individual applications waiting far longer. The delay produced a durable political argument known as the drug lag. The argument held that therapies available in other industrial countries reached American patients later because the American reviewer was understaffed rather than because the American standard was more demanding, and patient advocates pressing for faster access to treatments for HIV infection and other serious diseases gave the argument moral force before 1992. The Food and Drug Administration attributed the backlog to a mismatch between statutory responsibility and resources. Review work had grown more demanding as applications grew larger and more complex, while appropriations had not kept pace, and Congress showed little willingness in 1992 to close the gap from general revenues alone.

The Prescription Drug User Fee Act of 1992, Public Law 102-571, resolved the impasse by charging the beneficiaries of faster review for the added capacity. The logic resembled a fee for service rather than a tax. A sponsor received a defined service, namely a review conducted by a staffed and equipped agency within a predictable period, and the fee paid for the reviewers, support staff, and information systems that made the service possible. Collections began in fiscal 1993, and the Congressional Research Service recorded that most of the early collections went toward hiring additional reviewers. The standard application fee in fiscal 1993 stood at approximately 100,000 dollars. Fee waivers and exemptions softened the charge for small businesses and for products aimed at orphan diseases or unmet public health needs, preserving the principle that the fee purchased review capacity rather than permission to market a product.

The Original Bargain on Appropriations

The political price of industry money was a set of guarantees designed to keep the fees from displacing the appropriations they were meant to enlarge. Fees were to supplement congressional appropriations, never to supplant them, and the statute enforced the promise through mechanisms that lawyers for the agency and for industry came to call triggers. Under the trigger structure, the authority to collect and spend fees operated only while direct appropriations for the human drug review program, and for the agency overall, remained at or beyond a floor tied to the pre 1992 funding level and adjusted for inflation under a statutory formula. A Congress that tried to substitute fee revenue for appropriated funds would switch the fee authority off and leave the review program with neither source at full strength. The trigger therefore protected the industry, which had agreed to pay for added capacity rather than for the baseline cost of government, and it protected the appropriations committees, whose role the fee program might otherwise have eroded.

A second feature of the 1992 settlement shaped everything that followed. Congress declined to write detailed review deadlines into the statute itself. The findings in the 1992 Act instead pointed to letters from Commissioner of Food and Drugs David Kessler, dated September 14, 1992 and September 21, 1992, addressed to the chairmen of the committees of jurisdiction, in which the agency committed itself to performance goals for the use of the fee revenue. The commitment letter device became the constitutional habit of the program. Statute supplies the money and the legal authority, while a negotiated letter supplies the operational promises, and each renewal since 1992 has reproduced the pattern at a larger scale.

How often does Congress have to renew FDA user fee authority?

Congress must renew the authority every five years. Each renewal sets fee revenue for a new five year period, revises the fee structure, and adopts a fresh set of performance commitments negotiated between the Food and Drug Administration and regulated industry before enactment.

The five year cycle functions as the metronome of the entire approval system. The first renewal arrived in the Food and Drug Administration Modernization Act of 1997, which tightened review timelines, including a ten month standard goal, and codified the fast track designation for drugs treating serious conditions. The second renewal rode inside Title V of the Public Health Security and Bioterrorism Preparedness and Response Act of 2002, a vehicle shaped by the anthrax attacks of 2001. The third renewal formed Title I of the Food and Drug Administration Amendments Act of 2007, enacted after controversies over the safety of marketed drugs, and that statute paired the money with new authority for risk evaluation and mitigation strategies and a strengthened postmarket safety apparatus. The fourth renewal, the Food and Drug Administration Safety and Innovation Act of 2012, created the breakthrough therapy designation and supply chain provisions alongside the fee provisions. The fifth renewal, the Food and Drug Administration Reauthorization Act of 2017, gathered the medical product fee programs into a single statute and rebuilt the prescription drug fee structure itself. The sixth renewal closed on September 30, 2022, when the Food and Drug Administration User Fee Reauthorization Act of 2022 carried the program through fiscal 2027, and the Consolidated Appropriations Act, 2023, enacted in December 2022, then added a further package of agency policy reforms on top of the renewed fee authority.

Each renewal demonstrates why the cycle matters beyond accounting. Because the fee authority expires, no participant can treat the settlement as permanent, and the approaching deadline forces the agency, the industry, patient organizations, and the committees of jurisdiction into a structured renegotiation on a predictable schedule. Policy riders accumulate in the renewal vehicles because the vehicles must pass. Fast track in 1997, risk evaluation and mitigation strategies in 2007, and breakthrough therapy in 2012 all traveled into law attached to fee bills that Congress could not easily allow to fail. The negotiation behind each package begins years before the deadline. The agency holds public meetings, consults patient and consumer representatives, and then negotiates directly with the regulated industry over fee revenue levels and the performance commitments the revenue will buy. The resulting commitment letter travels to Congress as the operational core of the package, and Congress legislates around it, accepting most terms while adding statutory policy of congressional choosing. Public and patient participation, modest in the earliest cycles, grew more formal in later cycles, with the 2012 statute strengthening patient involvement in drug development discussions.

What Sponsors Pay and How the Fee Burden Is Divided

The original 1992 design divided fee revenue into three equal parts. Application fees, paid by a sponsor for each new drug application or biologics license application submitted, supplied one third. Annual establishment fees, paid for each manufacturing establishment, supplied a second third. Annual product fees, paid for each covered product on the market, supplied the final third. The annual components stabilized revenue in years when submissions dipped.

The Food and Drug Administration Reauthorization Act of 2017 replaced that architecture with a simpler and more predictable one. The 2017 statute abolished the establishment fee and the product fee, eliminated the separate fee for supplemental applications, and substituted a single annual program fee assessed on sponsors for each prescription drug product identified in an approved application, subject to statutory limits and waivers. Under the 2017 structure, application fees supply roughly 20 percent of total fee revenue and program fees supply roughly 80 percent. Agency leaders defended the redesign on grounds of stability. Janet Woodcock, director of the Center for Drug Evaluation and Research, testified to the House Energy and Commerce Committee in March 2017 that the revised structure would increase predictability, stability, and efficiency in fee setting and would improve the ability of the agency to engage in long term financial planning. A revenue base weighted toward annual charges on approved products fluctuates less than a base weighted toward submissions, and a reviewer workforce hired against stable revenue can be retained across the application cycle.

Two sibling programs joined the prescription drug program in 2012 and share the five year rhythm. The Generic Drug User Fee Amendments, created by the 2012 statute, charged the generic industry for the review of abbreviated new drug applications and for inspections and program work that a large generic backlog had made urgent, with first collections in fiscal 2013. The Biosimilar User Fee Act, created by the same statute, financed the young pathway for biosimilar biological products through development fees and application fees, also with first collections in fiscal 2013. Both programs adopted the family pattern of negotiated commitments, published goals, and annual financial reporting, so that by the 2017 and 2022 rounds Congress was reauthorizing a portfolio of fee programs in a single negotiation.

The Growth of the Fee Share

The dependence created in 1992 deepened steadily, and the honest way to describe the fee share is as a trajectory anchored to named years rather than as a single timeless percentage. The Congressional Research Service report titled FDA Human Medical Product User Fee Programs recorded that prescription drug user fees covered 7 percent of the total costs of the prescription drug program in fiscal 1993, with appropriations covering the remaining 93 percent. The supplement of 1992 began as a small addition to an appropriated program. Congressional Research Service report RL33914 recorded that user fees contributed 48.4 percent of the budget of the human drug program in fiscal 2008, approaching half of a program that Congress had once funded almost entirely from the Treasury. Congressional Research Service report R44750 recorded that user fees covered 75 percent of the total costs of the prescription drug program in fiscal 2023, with appropriations covering 25 percent.

The three figures measure slightly different denominators across three decades, and the Congressional Research Service analysis treats them as points on a single rising line rather than as interchangeable statistics. The direction is unmistakable under any consistent accounting. A funding source designed as a supplement became, across six authorizations, the majority source for the premarket review enterprise. The review workforce that fees built could not have been assembled from the appropriations of the period, and the same workforce would face dismantling if the fee stream stopped.

Performance Goals and the Meaning of Action

The consideration that sponsors receive for the fees is expressed as performance goals, and precision about the goals matters because loose language about deadlines misdescribes the law. In the mature form of the program, the commitment letters set a goal that the agency will review and act on 90 percent of standard new drug applications within ten months of the filing date and 90 percent of priority applications within six months, with priority status reserved for drugs offering significant improvement against serious conditions. The goals extend well beyond the final decision. First cycle goals govern the completeness and timeliness of the initial review round, and meeting management goals set time frames for scheduling the meetings through which sponsors and reviewers align on trial design and evidence, for issuing minutes after those meetings, for responding to requests for special protocol assessments, and for acting on clinical holds. Later commitment letters added goals for communicating anticipated postmarketing requirements before the action date and for labeling negotiations.

Three qualifications define what the goals legally are. The goals are commitments recorded in letters transmitted to Congress, not deadlines written into the operative text of the Federal Food, Drug, and Cosmetic Act, and no private right or automatic consequence follows from a missed goal date for a single application. The sanction for systematic failure is political and financial rather than judicial, because an agency that misses its goals weakens the case for renewal and for the fee levels the agency seeks. The goals also promise action rather than approval. An action can take the form of an approval letter, or of a complete response letter explaining why the application cannot be approved in the form submitted and identifying the deficiencies the sponsor must cure. The Pharmaceutical Research and Manufacturers of America made the distinction explicit in testimony at a public meeting in November 2005, stating that the metrics enacted in 1992 concerned only the timing of reviews and carried no assumption about the approvability of any application. A sponsor therefore purchases a decision on a schedule, and the purchase confers no claim on the content of the decision.

Dependence, Deadline Pressure, and the Capture Debate

The strongest criticism of the fee structure operates at two levels, one institutional and one empirical, and the institutional level came first. Sidney Wolfe, director of the Health Research Group at Public Citizen, argued in a statement issued on January 11, 2007 that the crucial drug regulatory functions of the agency were too important to be compromised by direct funding from the companies whose drugs the agency reviewed, and Public Citizen maintained that the entire cost of review should return to the Treasury through ordinary appropriations, as funding had run from 1906 through 1992. Congressional Research Service report RL33914 recorded the related concern in neutral terms, noting that dependence on fee revenue for staff retention might undercut congressional support for increases in direct appropriations. A 2006 survey by the Union of Concerned Scientists added testimony from inside the building, reporting that agency scientists described pressure connected to review timelines.

The empirical level arrived with Daniel Carpenter of Harvard University and coauthors Emily Zucker and Jerry Avorn, whose study titled Drug Review Deadlines and Safety Problems appeared in the New England Journal of Medicine in 2008. Carpenter and coauthors examined new molecular entities approved between 1950 and 2005, concentrating on applications submitted after the fee deadlines took effect in January 1993. The study found that approval decisions clustered in the weeks immediately before goal dates, and that drugs approved in the two months before a deadline carried higher odds of later withdrawal for safety reasons, of later receiving a boxed warning, and of later having a dosage form discontinued than drugs approved at other points in the cycle. The interpretation Carpenter offered was organizational rather than accusatory. A hard goal date concentrates deliberation at the end of a review, compresses the hours available for resolving late appearing safety questions, and rewards closure over continued inquiry, so a well intentioned workforce can produce systematically different decisions in the shadow of the clock.

The defense of the program answers at the same two levels. Institutionally, defenders including agency leadership under Commissioner Margaret Hamburg and industry associations such as the Pharmaceutical Research and Manufacturers of America pointed to the structure of the bargain itself. Fees flow under statute, at rates set through appropriations, toward goals published for public inspection, with annual performance and financial reports to Congress supplying an audit trail. Predictability, in the defense account, is the public benefit purchased by the arrangement. Sponsors plan development programs around known goal dates, patients gain earlier access when reviews finish months sooner, and the alternative on offer in 1992 and in each renewal debate was not a generously appropriated agency but the understaffed agency that produced the backlog, because Congress demonstrated across decades a reluctance to fund review expansion from general revenues at the scale the fees achieved. The agency also contested the empirical record directly. After the 2008 study appeared, agency analysts stated that agency records showed no statistically significant difference in safety outcomes between drugs approved near deadlines and drugs approved at other times, and Carpenter, after correcting errors the agency identified and extending the data, reported that the association persisted in the revised analysis.

The capture claim also resists clean testing. Deadline bunching is documented, and the association between pre deadline approval and later safety actions appeared in the 2008 analysis and in the corrected reanalysis, yet association confronts confounding at every step. Drugs approved near goal dates differ from other drugs in ways that statistics capture imperfectly, including product novelty and therapeutic urgency. Postmarket safety actions also reflect the surveillance system that detects them, and detection grew stronger across the decades the fee program matured, so a rising count of warnings can signal better detection rather than worse decisions. The counterfactual at the heart of the debate, namely the safety record of an equally resourced agency funded entirely by appropriations and governed by no goal dates, has never existed for observation. The evidence available through 2023 therefore supported a documented description and a serious argument on each side, without delivering a verdict that either side was entitled to claim as settled.

What Happens When Renewal Fails

The sunset is the enforcement mechanism behind the politeness of the negotiation. Fee authority under each authorization expires on a date certain, October 1 of the renewal year in the modern pattern, and expiration ends the legal power to collect the fees on which the review workforce depends. The Government Accountability Office examined the mechanics in the first renewal crisis of 1997 and reported that staff funded by fees would face a reduction in force if the authority lapsed, with federal personnel rules requiring notice sixty days before a reduction took effect. Notices therefore come due in the summer of a renewal year, weeks before the statutory deadline, and the prospect of issuing them concentrates congressional attention. Carryover balances can finance an orderly wind down or a brief continuation, as the 1997 analysis explained, but carryover cannot sustain a review enterprise in which fees covered 75 percent of program costs in fiscal 2023.

The discipline runs in both directions, which is why the cycle has never in fact been allowed to lapse into a shutdown of the review program. The industry faces the loss of the predictable review system the fees purchase, including goal dates around which launch planning and financing are organized. The agency faces the loss of the workforce that the fees employ and the return of the backlog conditions that discredited the pre 1992 system. Congress faces accountability for a visible and attributable failure. The 2022 renewal illustrated the pressure in compressed form, with enactment arriving on September 30, 2022, the final day of the expiring authority. A funding device adopted in 1992 as an experiment thus hardened into the load bearing wall of approval law, renewed six times because every party feared the consequences of removal more than the costs of renewal.

The Expedited Pathways and the Serious Disease Frame

The standard approval clock made sense for ordinary conditions and strained badly for fatal ones. A sponsor gathered evidence through phased trials, assembled a new drug application or a biologics license application, and waited while the Food and Drug Administration tested the submission against the statutory standard of safety and substantial evidence of effectiveness. That sequence protected patients from unproved products, and the protection carried a cost measured in time. For a slowly progressing condition with effective therapy available, the cost could be defended as the price of certainty. For a condition that killed or irreversibly disabled patients while no satisfactory therapy existed, the same delay read differently to Congress, to the agency, and to patients who organized during the acquired immunodeficiency syndrome epidemic of the 1980s. The expedited programs grew from that collision, preserving the approval statute while carving faster routes through it.

The serious and life threatening frame supplied an eligibility filter that limited every exception. Agency guidance on expedited programs for serious conditions treated a condition as serious when morbidity had substantial impact on day to day functioning, when the condition was persistent or recurrent, or when progression to death or irreversible disability was plausible without intervention. A condition qualified as life threatening when death was likely within a matter of months or when premature death was likely without early treatment. No program operated as a general fast lane for commercial convenience. Priority review, fast track, breakthrough therapy, regenerative medicine advanced therapy, and accelerated approval all required a sponsor to cross the seriousness threshold first. Only after that crossing did the programs diverge over the additional showing required and the benefit conferred.

The divergence produced a layered system rather than a single alternative pathway. Some programs changed only the pace of agency action after an application arrived. Some changed the intensity of contact during development, long before an application existed. Accelerated approval changed the endpoint on which approval itself could rest, while attaching a post approval obligation to confirm predicted benefit. A sponsor could hold several designations for one product at once, because each designation answered a different bottleneck. A product could receive fast track designation during development, earn breakthrough therapy designation after early clinical results appeared, proceed on a surrogate endpoint under accelerated approval, and then receive priority review of the completed application. The stacking reflected a legislative judgment that delay had distinct causes requiring distinct tools.

Priority Review and the Shortened Clock

Priority review addressed the queue for agency action on a completed application. The designation originated in agency policy adopted under the Prescription Drug User Fee Act of 1992, under which the agency agreed to performance goals and created a two tier system of standard review and priority review. The goal for action on a standard application stood at ten months. The goal for a priority application stood at six months. Designation directed attention and resources toward products that, if approved, would provide a significant improvement in safety or effectiveness for treatment, diagnosis, or prevention of a serious condition. Increased effectiveness, elimination or substantial reduction of a treatment limiting reaction, enhanced compliance expected to improve serious outcomes, or evidence in a new subpopulation could support that judgment, according to agency descriptions of the program.

Priority review changed the clock and changed nothing else. Agency statements stressed that designation did not alter the scientific and medical standard for approval and did not lower the quality or quantity of evidence required. The designation did not shorten the trial period before submission, because the six month goal ran from filing rather than from the start of development. A sponsor whose trials required five years gained no reduction in those years from priority review alone, which explained why Congress later built programs that intervened during development. Priority review remained an agency determination for every application, although a sponsor could expressly request designation. The classification functioned as triage expressed through the calendar rather than through relaxation of proof.

A later invention turned the shortened clock into a transferable reward. Congress created the first priority review voucher program in the Food and Drug Administration Amendments Act of 2007 for neglected tropical diseases, where weak purchasing power supplied little commercial incentive. The Food and Drug Administration Safety and Innovation Act of 2012 added a voucher program for rare pediatric diseases, and legislation in 2016 added a program for medical countermeasures. Approval of a qualifying product earned a voucher entitling the holder to priority review of a different application that might not have qualified alone, and the voucher could be transferred or sold. The Government Accountability Office recorded agency concern in a 2016 report that voucher reviews could strain workload and displace agency ranking by clinical importance.

What does fast track designation actually do?

Fast track designation obliges the Food and Drug Administration to facilitate development and expedite review for a drug intended for a serious condition that shows potential to address an unmet medical need, through more frequent communication and review of completed application sections before the full application arrives.

Congress created the program in section 112 of the Food and Drug Administration Modernization Act of 1997, which added section 506 to the Federal Food, Drug, and Cosmetic Act, codified for fast track at 21 U.S.C. 356(b). The statute directed the agency, at sponsor request, to facilitate development and expedite review for a product intended for a serious or life threatening condition that demonstrated potential to address an unmet medical need. Seriousness alone did not suffice. A serious disease with several effective therapies and no gap in care offered a weak case, while a serious disease with no approved therapy offered a strong case. Superior effectiveness, avoidance of serious side effects, improved diagnosis where earlier diagnosis improved outcomes, or reduced toxicity could show that potential.

The timing of designation separated fast track from priority review. A request could rest on preclinical data, on early clinical data, or on both, because the statute asked for demonstrated potential rather than demonstrated improvement. A sponsor did not have to wait for a pivotal trial, or even for proof in patients, and a request could accompany the investigational new drug application or follow afterward. The agency undertook to decide within sixty days. The benefits operated mainly during development, through more frequent meetings on the development plan and more frequent written communication on trial design and biomarker use. Disagreement over endpoints or dose discovered after a large trial enrolled patients wasted years. The same disagreement resolved at design stage cost weeks, and fast track made early resolution a statutory entitlement.

Rolling review was the second signature benefit. Under ordinary practice, review began after submission of the complete application, so a defect in an early section surfaced only after every section had been assembled. A fast track sponsor could submit completed sections for review before the remaining sections were finished, and the agency could begin work in advance. Rolling review did not guarantee approval, and designation did not guarantee accelerated approval or priority review, although a fast track product was eligible for consideration for both when separate criteria were met. Designation could be withdrawn if later data no longer supported the criteria. The sponsor gained access, sequencing, and speed of feedback, while the final approval decision demanded the same substantial evidence as the standard pathway.

Breakthrough Therapy and Intensive Guidance

Breakthrough therapy designation raised the entry bar and raised agency involvement in return. Congress created the designation in section 902 of the Food and Drug Administration Safety and Innovation Act of 2012, placing breakthrough therapy in section 506(a) of the Federal Food, Drug, and Cosmetic Act, codified at 21 U.S.C. 356(a). The product had to be intended, alone or in combination, to treat a serious or life threatening condition. Preliminary clinical evidence also had to indicate that the product may demonstrate substantial improvement over existing therapies on one or more clinically significant endpoints, such as substantial effects observed early in development. Preliminary meant the evidence need not establish effectiveness for approval. Clinical excluded a case built on preclinical data alone. Substantial set a threshold above incremental gain.

The benefits committed senior agency resources to the development program. The statute and agency implementation provided intensive guidance on efficient development and involvement of senior managers and experienced review staff in a collaborative, cross disciplinary review. Agency procedure assigned a cross disciplinary project lead as scientific liaison between review team and sponsor, provided timely advice and interactive communication, and directed steps to make trial design as efficient as practicable, including minimizing patient exposure to a potentially less efficacious treatment when scientific considerations allowed. Sponsors also received fast track benefits, including rolling review. Fast track increased frequency of contact, while breakthrough therapy changed the seniority and structure of contact, on the theory that a large early effect presented design questions requiring decision makers rather than correspondents.

The legal difference between the designations was concrete, despite confusion in commercial reporting that treated both labels as generic regulatory favor. Fast track asked whether nonclinical or clinical data showed potential to fill an unmet need. Breakthrough therapy asked whether preliminary clinical evidence showed a prospect of substantial improvement over available therapy on a clinically significant endpoint. A product could hold fast track designation without approaching the breakthrough threshold, and a product could gain breakthrough designation after fast track designation once human data matured. Agency decisions were due within sixty days, and agency practice encouraged submission no later than the end of phase two meeting, because guidance after pivotal design arrived too late to change that design. Designation promised no outcome, and early effects in a small population could disappear in larger trials.

Regenerative Medicine Advanced Therapy in Brief

Section 3033 of the 21st Century Cures Act of 2016 added regenerative medicine advanced therapy designation in section 506(g) of the Federal Food, Drug, and Cosmetic Act. A qualifying product had to meet the definition of a regenerative medicine therapy, covering cell therapies, therapeutic tissue engineering products, human cell and tissue products, and combination products using such therapies, with certain human gene therapies within agency interpretation of the definition. The product had to be intended to treat, modify, reverse, or cure a serious or life threatening condition, and preliminary clinical evidence had to indicate potential to address unmet medical needs. Benefits tracked breakthrough therapy, with intensive early interaction and eligibility for priority review and accelerated approval when separate criteria were met.

Accelerated Approval and the Surrogate Endpoint Bargain

Accelerated approval changed the basis for approval rather than the pace or staffing of review. The program began as regulation. In 1992 the agency issued rules for drugs in 21 C.F.R. Part 314, Subpart H, and for biologics in 21 C.F.R. Part 601, Subpart E. The rules responded to pressure intensified during the acquired immunodeficiency syndrome crisis, when waiting for proof of extended survival meant withholding a product from patients whose survival was measured in months, even after a strong effect on a disease marker had appeared. Congress codified and expanded the program in section 901 of the Food and Drug Administration Safety and Innovation Act of 2012, amending section 506(c), codified at 21 U.S.C. 356(c), and directing the agency to consider the availability of alternative treatments.

The mechanism turned on endpoint distinctions. A clinical endpoint measured how a patient felt, functioned, or survived, and proof on such an endpoint could require long follow up in a large population. A surrogate endpoint was a laboratory measurement, radiographic image, physical sign, or other marker that was not itself clinical benefit but could stand in for such benefit. An intermediate clinical endpoint was a therapeutic effect observable earlier than irreversible morbidity or mortality. Accelerated approval allowed approval for a serious condition on an effect on a surrogate or intermediate clinical endpoint reasonably likely to predict clinical benefit. Reasonably likely sat below the validation demanded for traditional approval. The agency weighed scientific support for the endpoint, magnitude and duration of effect, disease seriousness, and available therapy, and supporting studies still had to be adequate and well controlled.

Three conditions bounded that discretion. The disease had to be serious. The product had to offer a meaningful advantage over available therapy, expressed in regulation as meaningful therapeutic benefit over existing treatments and carried forward through attention to unmet need and availability of alternatives. A product duplicating effective therapy without advantage in effectiveness, safety, tolerability, or service to an unserved population did not fit, even when a convenient surrogate existed. The third condition was confirmatory trials. Approval rested on a prediction, and the statute required adequate and well controlled post approval studies to verify and describe predicted clinical benefit, conducted with due diligence on an agency schedule with milestones and reporting. Labeling had to disclose the limits of the evidence, so prescribing reflected the distinction between predicted and verified benefit.

The confirmatory requirement supplied the enforcement logic of the bargain. When a required study failed to verify benefit, or a sponsor failed to conduct a study with due diligence, the program provided for withdrawal of approval under regulatory and statutory procedures. Withdrawal in practice was slow and contested, and products that remained marketed while confirmatory evidence lagged became a subject of congressional oversight and Government Accountability Office reporting on postmarket studies. The difficulty did not change the legal structure, because verification was a condition written into the approval itself, accepted in exchange for marketing years earlier than traditional proof allowed. Oncology and human immunodeficiency virus treatment became the heaviest uses in agency practice, since tumor response and viral load could be measured far earlier than survival.

A persistent misconception requires careful correction. Accelerated approval is a full approval subject to conditions, not a provisional license that expires on its own. A product receiving accelerated approval entered the market as an approved product, could be prescribed for the approved indication, and remained approved without any automatic sunset or conversion event. Nothing in the statute or the 1992 regulations set a term after which approval lapsed if confirmation remained incomplete. Conditions operated through obligation and enforcement rather than expiration. The sponsor owed confirmatory studies with due diligence, the agency retained authority to pursue withdrawal after failure to verify benefit or failure of diligence, and traditional approval followed when confirmatory evidence verified benefit and the agency removed the conditions. Calling the status temporary approval understates market rights conferred and overstates automatic patient protection, because continued marketing depended on agency action rather than a self executing deadline.

Rare Diseases, Orphan Products, and the Outside Path

Rare disease products leaned on expedited programs more heavily than products for common conditions, for reasons of arithmetic. A small population produced small trials, long recruitment, and endpoints difficult to validate in advance, while patients often faced serious illness without approved therapy. A sponsor had strong reason to seek fast track communication early, pursue breakthrough designation when preliminary data showed a large effect in a small cohort, and propose a surrogate endpoint for accelerated approval when a trial powered for survival would require more patients than the population could supply. The incentive layer making such development commercially plausible sat outside the expedited statutes, in the orphan drug incentive structure, combining marketing exclusivity, tax credits for clinical testing, and waiver of application fees. Orphan incentives improved return on development, while expedited programs reduced time and evidentiary friction, and neither layer substituted for the other.

The right to try legislation of 2018 marked the boundary of the system by creating a route outside approval altogether. The Trickett Wendler, Frank Mongiello, Jordan McLinn, and Matthew Bellina Right to Try Act, Public Law 115-176, signed on May 30, 2018, permitted an eligible patient to seek an eligible investigational drug directly from a sponsor or manufacturer without agency authorization of individual use. Patient eligibility required a life threatening condition, exhaustion of approved options, inability to participate in a trial, written informed consent, and physician certification. Product eligibility required completion of a phase one trial, an active investigational program, and continued investigation toward approval. The statute imposed no duty on a sponsor to provide a product, limited liability for compliant provision or refusal, restricted agency use of outcomes in approval decisions except in defined circumstances, and required annual sponsor reporting.

The contrast was the agency expanded access program, often called compassionate use, which the right to try law left in place. Expanded access allowed a patient with a serious or immediately life threatening condition, without satisfactory alternatives and without trial access, to receive an investigational product under agency authorization, with institutional review board oversight, informed consent, sponsor agreement, and agency authority to impose a clinical hold for safety concerns. A Government Accountability Office report in 2017 found that the agency had allowed nearly all expanded access requests examined to proceed. Expanded access kept individual use inside the regulated investigational system, while right to try removed agency pre authorization and institutional review board approval as prerequisites and placed the decision among patient, physician, and sponsor. By August 15, 2018, the agency had published implementation information, and the two systems operated in parallel for patients whose disease allowed less time than ordinary proof required.

Post-Marketing Obligations and the Withdrawal Problem

Approval closed one proceeding and opened another. A drug that survived review entered the market with a record built from controlled trials in selected populations under protocol conditions. Marketed use was broader, longer, and less controlled. Patients were older, sicker, pregnant, taking other medicines, or using the product in ways the trials had not tested. The statutory design treated that gap as a reason for continuing obligations. The sponsor that won approval acquired duties to keep studying the product, to report what happened to patients who took the product, to operate special controls for particular risks, and to keep the labeling aligned with evidence gathered after launch. The withdrawal machinery in that design, for much of the period covered in this article, moved far more slowly than the obligations accumulated, and the distance between a duty to generate evidence and a power to act on evidence became the central post-marketing problem in FDA approval law.

Required Studies and the Difference a Requirement Makes

The modern authority for compelled post-marketing study came from the Food and Drug Administration Amendments Act of 2007, which added section 505(o)(3) to the Federal Food, Drug, and Cosmetic Act, codified at 21 U.S.C. 355(o)(3). The provision authorized the Food and Drug Administration to require post-marketing studies or clinical trials at the time of approval or after approval, including when the agency became aware of new safety information. A required study could assess a known serious risk related to use of the drug, assess signals of a serious risk, or identify an unexpected serious risk when available data indicated the potential for such a risk. The authority mattered because it replaced reliance on negotiation alone. A requirement under section 505(o)(3) carried a timetable, periodic reporting, and consequences for failure to proceed with due diligence absent good cause.

The distinction between a post-marketing requirement and a post-marketing commitment organized the whole system, and confusion between the two terms concealed where legal force actually lay. A requirement was a study or trial the sponsor had to conduct because a statute or regulation imposed the obligation. Requirements arose under several authorities, including section 505(o)(3) for safety studies, the accelerated approval provisions for confirmatory trials, and pediatric study provisions in other parts of the Act. A commitment, by contrast, was a study or trial the sponsor agreed in writing to conduct and that did not arise from a statutory or regulatory mandate, so the source of the obligation was agreement rather than command and enforcement differed accordingly.

Both categories ran on a reporting clock. Section 505(o)(3) required the sponsor to report periodically on the status of each required study or trial, and the implementing practice folded that duty into the annual status report structure built under section 506B of the Act and the regulations at 21 C.F.R. 314.81 and 21 C.F.R. 601.70. The annual report described each open requirement or commitment, stated the schedule for completion, characterized the status in categories the agency defined, and explained delays. The Food and Drug Administration in turn published reports on the performance of drug and biologic sponsors in conducting post-marketing requirements and commitments, organized by fiscal year. The annual rhythm also revealed the weakness of the instrument. A status report could document slippage year after year without, by itself, completing a trial or removing a product, and the history of confirmatory trials under accelerated approval showed how long documented delay could persist before any further consequence followed.

Adverse Event Reporting and Signal Detection

The second post-marketing system collected experience rather than testing hypotheses. Under 21 C.F.R. 314.80, the holder of an approved application had to report each adverse drug experience that was both serious and unexpected as soon as possible and in no case later than 15 calendar days after initial receipt of the information. The regulation required prompt investigation of reported experiences and follow up reports when new information arrived, and the duty attached regardless of source, whether a report came from a health professional, a patient, the literature, or a foreign source. Section 314.81 supplied the companion duties, including field alert reports for problems with distributed product and the annual report summarizing significant new information that might affect safety, effectiveness, or labeling. Together the two regulations made the sponsor the first collector of the nation’s experience with the sponsor’s own product, giving the company with the deepest knowledge of the product a legal duty to pass alarming knowledge to the regulator on a fixed clock.

Those reports flowed into the agency adverse event reporting database, known during the period covered here as the Food and Drug Administration Adverse Event Reporting System. The database held reports submitted by sponsors, reports submitted directly by clinicians and patients through the MedWatch program, and reports drawn from published literature. Agency safety staff screened the database for signals, meaning patterns that suggested a causal relationship between a product and an adverse outcome that labeling did not describe or described too weakly. Signal detection was an institutional function with defined limits. A spontaneous report was not proof of causation. The work therefore combined case review, statistical screening, literature review, and, when a signal held up, referral for further study or for action through labeling, a risk strategy, or a required study under section 505(o)(3). The 2007 amendments strengthened this side of the house as well, directing regular screening of the database and quarterly public posting of new safety information and potential signals of serious risk. The system generated suspicion at scale. Turning suspicion into a regulatory decision remained a separate and slower process.

Risk Evaluation and Mitigation Strategies

For a defined set of products, Congress concluded in 2007 that labeling and reporting were not enough. The same amendments act that created section 505(o)(3) added section 505-1 to the Act, codified at 21 U.S.C. 355-1, authorizing risk evaluation and mitigation strategies. A strategy was required when the agency determined that such a strategy was necessary to ensure that the benefits of a drug outweighed its risks. The determination could occur before approval or after approval when the agency became aware of new safety information. The basic elements included a timetable for submission of assessments of the strategy, and, depending on the product, a Medication Guide, a patient package insert, or a communication plan directed at health care providers.

Elements to assure safe use could require that prescribers hold particular training, experience, or certification, that pharmacies or health care settings be specially certified to dispense the product, that dispensing occur only in specified settings, that dispensing depend on documentation of safe use conditions such as laboratory test results, that each patient be subject to monitoring, or that each patient be enrolled in a registry. A strategy containing such elements also required an implementation system through which the sponsor monitored and evaluated implementation and worked to improve it. The compliance burden was substantial and continuous. Assessments of a strategy were due by 18 months, by 3 years, and in the seventh year after approval of the strategy unless the agency specified a different schedule. The strategy thus converted a risk judgment made at approval into an operating program that had to be staffed and documented for as long as the strategy remained in effect, and modification or release of a strategy required agency action rather than a sponsor decision.

How Can a Manufacturer Strengthen a Warning Before the Agency Approves the Change?

Under the changes being effected regulation, the holder of an approved application may distribute a drug with a strengthened warning once the Food and Drug Administration receives the supplement. The change must rest on newly acquired information and meet the labeling evidence standard, and the agency may later disapprove it.

The provision behind that answer sat at 21 C.F.R. 314.70(c)(6)(iii) and governed one category of the changes being effected supplement. Most changes to an approved application required a supplement approved before distribution. The labeling provision worked faster by design. The regulation allowed the holder of an approved application to add or strengthen a contraindication, warning, precaution, or adverse reaction, to add or strengthen a statement about abuse, dependence, psychological effect, or overdosage, to strengthen a dosage instruction intended to increase safe use, or to delete a false, misleading, or unsupported indication or claim of effectiveness, and to begin distribution of the product bearing the revised labeling upon agency receipt of the supplement rather than after agency approval. The condition was that the change reflect newly acquired information, a term the regulations defined to include data, analyses, and other information not previously submitted, whether derived from clinical studies, adverse event reports, a post-marketing study, peer reviewed literature, or the scientific record more generally. The evidence of a causal association also had to satisfy the standard for inclusion in the labeling under 21 C.F.R. 201.57.

Treated as a compliance mechanism, the supplement performed several functions inside the post-marketing system. The mechanism removed timing as an excuse. A brand manufacturer that acquired information supporting a stronger warning did not have to wait through a prior approval cycle before the revised warning could reach prescribers and patients, and the filing itself created a dated record of what the manufacturer knew, what wording the manufacturer chose, and when the manufacturer acted. The agency retained authority over the result. Reviewers could approve the supplement, request different wording, or disapprove the change and order distribution under the revised labeling to stop. Changes outside the enumerated categories followed the prior approval route instead. Within the post-marketing architecture, the changes being effected supplement was therefore the fast lane for one direction of travel, toward a stronger warning, and the existence of that lane shaped how promptly the labeling system could respond when adverse event reporting and post-marketing study produced new risk information.

The Withdrawal Machinery and Why the Machinery Ran Slowly

Withdrawal stood at the far end of the system, and the statute made withdrawal a formal proceeding rather than an administrative shortcut. Under 21 U.S.C. 355(e), the agency could withdraw approval on grounds that included a finding, after due notice and opportunity for hearing, that the drug was unsafe for use under the conditions prescribed in the labeling, that new evidence showed the drug was not shown to be safe, that new evidence showed the drug was not as effective as the approval record had represented, or that the application contained an untrue statement of material fact. The procedural words did much of the work. The agency issued a notice of opportunity for hearing stating the grounds for the proposed withdrawal. The sponsor could waive a hearing and accept the result, or the sponsor could demand a hearing and put the agency to its proof. When a sponsor demanded a hearing, the proceeding was built on a written record, with submissions, evidence, and argument resembling trial practice, and the final decision rested with the Commissioner, subject to judicial review in the courts of appeals.

The accelerated approval regulations added a parallel and nominally faster track. Under 21 C.F.R. 314.530 for drugs and 21 C.F.R. 601.43 for biologic products, the agency could withdraw an accelerated approval when a required confirmatory study failed to verify clinical benefit, when other evidence demonstrated that the product was not safe or effective under its conditions of use, when the sponsor failed to conduct a required study with due diligence, when restrictions on distribution proved inadequate, or when promotional material was false or misleading. A sponsor that contested withdrawal could still insist on process and submit extensive argument, and under the general withdrawal authority the sponsor could demand the hearing that section 355(e) promised. During the contest the product ordinarily remained on the market and remained prescribable, because a proposal to withdraw was not itself a withdrawal. The combination of a hearing right, a record built for judicial review, and continued marketing during the proceeding produced the characteristic delay. Each contested withdrawal consumed years and consumed senior agency attention out of proportion to the number of products involved, and resistance carried little downside for a sponsor beyond legal cost while approval and sales continued.

Bevacizumab and the First Contested Hearing

Bevacizumab, marketed as Avastin by Genentech, supplied the defining test of the machinery. On February 22, 2008, the Food and Drug Administration granted accelerated approval to bevacizumab in combination with paclitaxel for patients who had not received chemotherapy for metastatic HER2 negative breast cancer. The approval rested on a trial showing a substantial improvement in progression free survival, an endpoint the agency accepted as reasonably likely to predict clinical benefit, and the approval carried the standard accelerated approval condition that further trials verify the benefit. The confirmatory program did not reproduce the size of the original effect. Subsequent trials showed smaller gains in progression free survival, no improvement in overall survival, and a risk profile that included severe and in some cases fatal toxicities. In December 2010, after an advisory committee had recommended withdrawal, the agency proposed to withdraw the breast cancer indication and issued a notice of opportunity for hearing.

Genentech demanded a hearing, and the hearing that took place in June 2011 was the first contested hearing of its kind on a proposal to withdraw an accelerated approval indication. The agency and the sponsor presented evidence and argument before Commissioner Margaret Hamburg, and the Oncologic Drugs Advisory Committee voted 6 to 0 that the confirmatory trials had not verified clinical benefit and that the indication should be withdrawn. On November 18, 2011, Commissioner Hamburg issued a decision withdrawing approval of the breast cancer indication, concluding that the available evidence showed no benefit sufficient to outweigh the risks of serious adverse events. Because approval attached indication by indication, withdrawal operated on the same unit, and the agency could excise a failed use while leaving the product and its other indications undisturbed. The decision removed one use from the label of a product that remained approved for other cancers. From proposal in December 2010 to decision in November 2011, the contested path consumed eleven months even in a case the agency treated as a priority and in which every advisory vote ran the same way.

Hydroxyprogesterone Caproate and a Twelve Year Interval

Hydroxyprogesterone caproate, marketed as Makena, traced the same arc over a much longer span. The Food and Drug Administration granted accelerated approval on February 3, 2011, for reducing the risk of preterm birth in women with a singleton pregnancy who had a history of singleton spontaneous preterm birth. The approval rested on an intermediate endpoint, a reduction in deliveries before 37 weeks of gestation, drawn largely from a single earlier trial, and the approval required a post-marketing trial to verify and describe clinical benefit in the health of newborns. When PROLONG reported in 2019, the trial had failed to verify clinical benefit. The results showed no reduction in the composite outcome of neonatal morbidity and mortality and no treatment effect on the intermediate endpoint that had supported the original approval.

On October 5, 2020, the Center for Drug Evaluation and Research proposed withdrawal of approval for Makena and its generic equivalents and issued a notice of opportunity for hearing, stating two grounds drawn from the accelerated approval framework, that the confirmatory trial had failed to verify clinical benefit and that the available evidence did not show the product to be effective for its approved use. The sponsor, by then Covis Pharma following a chain of acquisitions, requested a hearing. The hearing followed in October 2022, two years after the proposal, and the Obstetrics, Reproductive and Urologic Drugs Advisory Committee voted 14 to 1 against allowing the product to remain marketed while another confirmatory study was designed and conducted. In March 2023 the sponsor announced agreement to withdraw the product voluntarily while seeking an orderly wind down period. On April 6, 2023, Commissioner Robert Califf and Chief Scientist Namandjé Bumpus issued the final decision withdrawing approval of Makena and its generic versions and declining the requested delay. Measured from approval in 2011 to withdrawal in 2023, the interval was twelve years, during most of which the product held approval and was administered to the population the failed trial had been meant to protect.

The Statutory Repair in the Food and Drug Omnibus Reform Act

Congress addressed the structure directly at the end of 2022. Section 3210 of the Food and Drug Omnibus Reform Act of 2022, titled as a modernization of accelerated approval, was enacted on December 29, 2022, as part of the Consolidated Appropriations Act, 2023, Public Law 117-328. The provision amended section 506(c) of the Act and attacked delay at three points, the start of the confirmatory trial, the visibility of progress during the trial, and the procedure for ending approval after a failed trial. At the front end, the amendment required the agency, no later than the date of accelerated approval, to specify the conditions for the required post-approval studies, including the design, milestones, and target completion dates. The amendment also authorized the agency to require, as appropriate, that a confirmatory study already be underway before approval was granted, or that the study begin within a specified period after approval.

During the life of the approval, the amendment imposed a reporting interval measured in months rather than years. A sponsor had to submit reports to the agency every 180 days on the progress of each required confirmatory study, and the agency was directed to publish information from those reports. At the back end, section 3210 codified expedited withdrawal procedures in the statute itself. The procedures required the agency to give the sponsor notice and an explanation of the proposed withdrawal, an opportunity to meet with the agency and to submit a written appeal, an opportunity for public comment on the proposal, and an advisory committee meeting if the sponsor requested one and no advisory committee had previously advised the agency on the matter. The design traded the open ended hearing model for a defined sequence of steps that preserved sponsor participation and a public record while removing the trial form that had made each contested withdrawal a multi year project.

An Enforcement Record of Rare Contested Withdrawals

Measured against the volume of accelerated approvals granted since the pathway was created in 1992, formal withdrawal after failed verification remained a rare outcome through the period ending with the 2022 amendments, and the pattern is best described qualitatively because published counts varied with the treatment of conversions and sponsor initiated removals. Many accelerated approval indications ended in quieter ways. A confirmatory program succeeded and the indication converted to regular approval. A sponsor withdrew an indication voluntarily, sometimes after discussions with the agency and sometimes after an advisory committee vote made the direction of the evidence plain, and voluntary withdrawal by sponsors accounted for a share of the exits from the pathway. Or an indication remained in a prolonged unresolved state, marketed under an approval whose confirmatory evidence was late, equivocal, or negative, while the agency weighed the cost of a contested proceeding. Oversight reports by the Government Accountability Office and the Office of Inspector General of the Department of Health and Human Services, published before the 2022 amendments, documented incomplete confirmatory trials and limits in agency follow through. The honest summary of the record is structural rather than numerical. The pre 2023 system made inaction cheap for sponsors and action expensive for the agency, and the case histories of bevacizumab and hydroxyprogesterone caproate showed the expense in concrete terms, eleven months of contested process for one indication in one case and twelve years from approval to withdrawal in the other. Section 3210 was written against that record, and the provisions on studies underway at approval, on reports every 180 days, and on codified expedited withdrawal were the legislative answer to a problem that post-marketing obligations alone had documented but had not solved.

Advisory Committees and the Preemption Asymmetry

Two features of implementation reveal more about FDA approval law than the approval standard alone reveals. The first is advice. Before the Food and Drug Administration approves, refuses, or withdraws a drug, the agency often places the hardest scientific questions before a committee of outside experts, in public, and then decides for itself. The second is liability. After approval, an injured patient may bring a claim under state law, and the fate of that claim can turn not on the injury or the molecule, but on whether the manufacturer held the original application or an abbreviated application. Advisory committees show how the agency borrows judgment without surrendering authority. Preemption doctrine shows how a labeling rule written for uniformity can allocate the right to compensation.

The Statutory Frame for Outside Advice

Advisory committees in drug review operate under two overlapping frames. The general frame is the Federal Advisory Committee Act, which governs advisory bodies across the federal government. That Act treats a committee established or used by an agency for advice as a public institution rather than a private consultation. A committee operates under a charter, its meetings ordinarily require advance public notice, its proceedings are open unless a lawful basis supports closure, and its records become part of the public record of agency decision making. The agency gains knowledge that resides outside government, in universities, hospitals, and clinical practice, and in exchange the agency accepts transparency about who advised the government, what material the advisers considered, and what the advisers recommended.

The specific frame sits in section 505(n) of the Federal Food, Drug, and Cosmetic Act, codified at 21 U.S.C. 355(n). That provision governs the use of advisory committees in the review of human drug applications and directs the agency toward referral where outside judgment carries the greatest weight, an application for a drug that contains a new molecular entity. Under section 505(n), such an application is expected to go before an advisory committee before approval, or the agency must explain why referral was unnecessary in the particular case. The expectation matters even when the agency departs from referral, because the departure requires a stated reason. In both paths, the statute pushes discretion toward explanation.

Membership completes the frame. Committees draw voting members principally from outside government, from physicians, statisticians, pharmacologists, and other specialists matched to the product under review, together in many drug committees with consumer and patient representatives. Because the experts most qualified to judge a new therapy often consult for manufacturers or hold financial interests in the same field, the agency screens prospective members for financial conflicts of interest. Screening can lead to exclusion, to recusal, or, where the governing standards allow and the agency judges the expertise necessary, to a waiver that permits participation despite a disclosed interest. Strict exclusion protects disinterested advice but can strip a panel of the few specialists who know a rare disease best. Liberal waiver practice preserves expertise but invites suspicion that advice was purchased in advance.

Advice That Binds No One

The legal character of a committee vote requires precise statement. A committee advises and the agency decides. No committee vote approves a drug, refuses a drug, or withdraws an indication. The Commissioner and the responsible center officials retain the statutory authority, and formal action issues in the name of the agency rather than the committee. The agency in practice follows committee recommendations in a substantial share of cases, since referral ordinarily signals genuine uncertainty and a lopsided vote from qualified outsiders carries institutional weight. Agreement, however, remains agreement rather than obedience, and disagreement remains lawful whenever the deciding official explains a different reading of the same evidence.

The withdrawal proceeding for the breast cancer indication of bevacizumab, marketed as Avastin, supplies the aligned case in a form the public record verifies in detail. The agency had granted accelerated approval for the breast cancer indication, and later confirmatory trials failed to verify the predicted benefit. The Center for Drug Evaluation and Research proposed withdrawal, the manufacturer Genentech contested the proposal and obtained a hearing, and the Oncologic Drugs Advisory Committee considered the matter during a two day hearing on June 28 and June 29, 2011. The committee voted six to zero in favor of withdrawing approval of the breast cancer indication. Commissioner Margaret A. Hamburg then issued the agency decision on November 18, 2011, revoking approval of that indication while leaving bevacizumab on the market for other approved uses. The vote supplied public expert judgment on a contested record. The revocation took legal effect through the Commissioner decision, grounded in a lengthy opinion on the hearing record, and not through the vote that preceded the decision by almost five months.

The institutional value of committees lies in three functions that a vote alone does not capture. A hearing makes the evidence contest public. Trial data that would otherwise move between sponsor and reviewers inside an application file become the subject of open presentations, agency critique, sponsor response, and testimony from patients and clinicians, stated where journalists and researchers can hear the disagreements directly. A committee process also forces the agency to state its questions in advance. The agency publishes briefing material and poses specific voting questions, and that framing identifies, before any vote, whether the dispute concerns efficacy, safety, trial design, or the population in which benefit might outweigh risk. A proceeding finally creates the record on which later litigation and journalism draw. Transcripts, briefing books, disclosed conflicts, and the stated basis for agency action remain available after the decision, so a court or a reporter begins from material generated in public rather than from recollection.

A Labeling Rule at the Center of Federalism

Preemption in prescription drug cases begins with an absence. The Federal Food, Drug, and Cosmetic Act contains no express preemption clause for prescription drugs. Congress wrote no provision for drugs comparable to the express clause for medical devices that the Supreme Court applied in Riegel v. Medtronic, 552 U.S. 312 (2008), to bar state requirements different from or in addition to federal device requirements. The device clause gave the Court in Riegel a textual instruction about displacement. The drug statute gives courts no comparable instruction. The doctrine for drugs therefore developed through impossibility preemption, which asks whether a party can comply with both federal law and state law at the same time. If simultaneous compliance is impossible, federal law prevails under the Supremacy Clause. If simultaneous compliance is possible, a state claim ordinarily survives an impossibility defense.

That test becomes concrete only after a court identifies who controls the operative federal duty. For failure to warn claims, the operative document is the approved label. Federal law requires marketing with approved labeling, and state tort law, as presented in the leading cases, imposes a duty to provide an adequate warning. Both bodies of law point at the same words on the same label. The decisive question is therefore not whether a warning matters, but who may change the label. A manufacturer with a lawful unilateral route to a stronger warning can satisfy a state duty without violating federal law. A manufacturer forbidden to depart from a federally fixed text cannot satisfy the same duty through the label at all. The decisions that follow turn entirely on that difference in permission, which traces back to the structure of original and abbreviated approval rather than to any legislative judgment about compensation.

Can a Generic Drug Claim Be Barred When a Brand Name Claim Is Not?

Yes. Under Mensing, a failure to warn claim against a generic manufacturer was held preempted because federal law required the generic label to match the brand label exactly, while under Levine a comparable claim against a brand manufacturer survived because the brand manufacturer could strengthen its warning without prior agency approval.

Wyeth v. Levine established the brand name side of the line. PLIVA, Inc. v. Mensing established the generic side. Mutual Pharmaceutical Co. v. Bartlett then extended the generic side from labeling to design. Read together, the decisions produce a rule that no single opinion announced as a general policy.

Wyeth v. Levine and the Brand Name Label

Diana Levine, treated for migraine and nausea, received the anti nausea drug Phenergan, manufactured by Wyeth, administered by the intravenous push method. The drug reached arterial blood, the exposure produced gangrene, and Levine lost an arm to amputation. Levine sued Wyeth in Vermont state court on the theory that the Phenergan label, although the label warned about risks associated with intravenous administration, failed adequately to warn against the danger that the intravenous push method presented. A Vermont jury found the label warning inadequate under state law and awarded damages, and the Vermont courts sustained the judgment. Wyeth contended that federal approval fixed the warning, that a state verdict requiring a stronger warning imposed a duty federal law made impossible to satisfy without agency permission, and that recognition of the claim would obstruct agency controlled labeling.

The Supreme Court rejected the defense in Wyeth v. Levine, 555 U.S. 555 (2009), by a vote of six to three. The majority rested first on the changes being effected regulation at 21 C.F.R. 314.70(c)(6)(iii). That regulation permits the holder of an approved new drug application, in defined circumstances that include strengthening a warning, to distribute a drug with a revised label upon filing a supplement, before the agency approves the revision, when the change rests on newly acquired information. That route defeated literal impossibility. Wyeth could have adopted a stronger warning about intravenous push administration and then sought agency review, so compliance with a state duty to warn more forcefully did not require violation of federal law. The majority rested second on responsibility. Through the history of federal drug regulation, the Court reasoned, the manufacturer has borne primary responsibility for its label, and agency approval of an initial label does not transfer that responsibility to the government. Approval sets a floor and a starting text. Post approval evidence and the changes being effected procedure place the continuing duty to keep the warning adequate principally on the firm that sells the drug.

The Court did not hold that every brand name claim escapes preemption. The opinion allowed that a claim could fail where clear evidence showed that the Food and Drug Administration would not have approved the stronger warning that state law allegedly required. On the record in Levine, Wyeth had not carried that showing, since argument about what the agency might have done did not equal evidence of rejection of the specific change. In a later decision, Merck Sharp and Dohme Corp. v. Albrecht in 2019, the Court held that the clear evidence question is a question of law for a judge rather than a question of fact for a jury. That refinement left the core Levine structure intact.

PLIVA v. Mensing and the Duty of Sameness

The plaintiffs in PLIVA, Inc. v. Mensing, 564 U.S. 604 (2011), took metoclopramide, the generic form of the brand name drug Reglan, over an extended period and developed tardive dyskinesia, a serious neurological disorder involving involuntary movements. The plaintiffs alleged that the labeling failed adequately to warn about the risk of developing that disorder from prolonged use, and the defendants were generic manufacturers rather than the holder of the original Reglan application. The change in defendant changed the federal rule governing the label. The generic pathway in the Hatch Waxman Amendments, codified in relevant part at 21 U.S.C. 355(j)(2)(A)(v), conditions abbreviated approval on a showing that the labeling proposed for the generic drug is the same as the labeling approved for the listed drug. The agency has long administered that sameness requirement as a continuing constraint that keeps the generic label aligned with the reference label, rather than as a snapshot satisfied on the approval date and then discarded.

Sameness removed the route that had saved the claim in Levine. A generic manufacturer could not use the changes being effected procedure to strengthen the warning unilaterally, because a strengthened generic label would no longer be the same as the brand label that federal law required the generic label to match. The generic manufacturers also could not be required to send unilateral warning letters to physicians that departed from the approved labeling, a route the Court rejected as inconsistent with the federal framework. The manufacturers could ask the agency to seek a change in the reference label, but a request dependent on agency agreement did not give a defendant the independent ability that impossibility analysis required. A duty that a party can satisfy only if a federal agency first agrees to move is not, under Mensing, a duty that the party can satisfy on its own while obeying federal law.

The Supreme Court held the claims preempted by a vote of five to four. Justice Thomas wrote for the majority, and the opinion confronted the distributional consequence directly. The majority acknowledged the apparent anomaly that a consumer remedy could turn on whether the pharmacy dispensed the brand name product or the generic product, even though the patient ingested the same drug and confronted the same warning. The majority treated that anomaly as a consequence of the statutory scheme Congress enacted, and therefore as a matter for Congress to alter. Justice Sotomayor wrote in dissent for four justices and challenged both the breadth of the impossibility holding and the allocation of responsibility the holding produced.

Bartlett and the Closing of the Design Route

A plaintiff whose warning claim fails under Mensing might turn to design, arguing that the drug itself was unreasonably designed. Mutual Pharmaceutical Co. v. Bartlett, 570 U.S. 472 (2013), closed that route for the generic drug before the Court. Karen Bartlett took a generic form of sulindac, an anti inflammatory drug, and developed a severe skin reaction that produced catastrophic injuries. Bartlett brought a design defect claim under New Hampshire law, and the Supreme Court held the claim preempted by a vote of five to four. A generic manufacturer could neither redesign the molecule, because federal law required equivalence to the reference drug and a different molecule would be a different drug outside the abbreviated approval, nor change the label, because Mensing had established that sameness foreclosed unilateral change. The majority also rejected the argument that a manufacturer escapes impossibility simply by ceasing to sell a lawful product. After Bartlett, the rule reached both classic theories, warning and design, for a manufacturer that could alter neither text nor molecule.

The Same Molecule in Two Legal Worlds

The claim that emerges from the three decisions can be stated plainly. The same molecule occupies two legal worlds. A patient injured after taking the brand name version may pursue a failure to warn claim that a jury decides on the adequacy of the warning, subject to the clear evidence qualification from Levine. A patient injured after taking the generic version of the identical drug, bearing a label with identical warning language, ordinarily finds the same claim preempted before any jury reaches adequacy. The patient ability to recover turns on a pharmacy substitution decision that the patient did not personally make. One manufacturer may change a label and the other may not. That single regulatory difference, the availability or unavailability of the changes being effected procedure, carries the entire weight of the asymmetry.

No part of that asymmetry rests on a congressional enactment that purports to decide who compensates injured patients. The approval statute was written about safety, effectiveness, and labeling control. The generic statute was written about equivalence, competition, and lower prices through abbreviated reliance on a reference drug. Neither statute was written as a liability statute, and neither contains a compensation scheme or a statement that consumers of generic drugs accept reduced tort rights as the price of reduced prices. The asymmetry is an emergent product of the two statutes read together with the labeling regulations and with impossibility doctrine. The Supreme Court asked, defendant by defendant, whether federal law left that defendant a unilateral path to a different warning or a different design. The answers differed because the regulatory positions differed, and tort law inherited the difference.

The asymmetry also feeds back into implementation. A brand name manufacturer faces jury exposure for an inadequate warning and therefore holds both a regulatory reason and a liability reason to use the changes being effected procedure promptly when new risk information emerges. A generic manufacturer after Mensing and Bartlett faces sharply reduced exposure for the same warning, and the principal engine for a generic label change becomes agency initiative or reference label revision rather than a unilateral supplement driven by litigation risk. Reporting duties and agency authority to require safety labeling changes continue to bind generic manufacturers, so the absence of tort exposure never meant the absence of regulation. Within the wider history of federal consumer protection law, the drug preemption cases stand as a clear modern demonstration that a rule about who may edit a document can decide, years later and in a different courtroom, who bears the cost of an injury that the document failed to prevent.

The Reform That Never Took Effect

The agency attempted to alter the premise on which Mensing rested, and the failure of that attempt completes the institutional story. In November 2013, the Food and Drug Administration published a proposed rule titled Supplemental Applications Proposing Labeling Changes for Approved Drugs and Biological Products, at 78 Federal Register 67985 on November 13, 2013. The proposal would have permitted holders of abbreviated new drug applications to use a changes being effected supplement to distribute revised safety labeling based on newly acquired information, in advance of agency review, even though the revised generic labeling would differ temporarily from the reference listed drug labeling. Had the proposal taken effect, a generic manufacturer would have possessed the independent capacity whose absence drove the result in Mensing, and impossibility analysis in later cases would have confronted a different regulation. The agency framed the proposal as a safety communication reform that would speed new risk information to prescribers and patients.

Generic manufacturers and the trade association then known as the Generic Pharmaceutical Association argued that independent labeling would undermine sameness, would produce divergent labels for the same drug during the interval before conformity, and would confuse prescribers and patients. Generic industry commenters also argued that the proposal would impose substantial pharmacovigilance costs on firms organized around equivalence, would raise prices, and would expose generic manufacturers to the liability that Mensing had held preempted. Public Citizen and plaintiffs lawyers who supported the proposal argued that a manufacturer selling a drug should bear responsibility for updating warnings about the drug. The agency under Commissioner Scott Gottlieb ultimately credited the sameness, confusion, and cost concerns. The agency delayed the proposal repeatedly, never issued a final rule, and withdrew the proposal in December 2018, with Commissioner Gottlieb and Center Director Janet Woodcock announcing the withdrawal on December 13, 2018. Public Citizen, through Sidney Wolfe of the Health Research Group, condemned the withdrawal as a failure to modernize generic labeling. The trade association for generic manufacturers praised the withdrawal. Congress enacted no statute reversing Mensing or otherwise allocating failure to warn liability between brand name and generic manufacturers.

The episode carries a lesson about the limits of administrative repair. The Supreme Court in Mensing had described the anomaly as a matter for Congress. The agency then attempted a regulatory repair that rested on contested authority over sameness, since a rule permitting even temporary divergence had to coexist with a statute that made sameness a condition of abbreviated approval. Delay accomplished what a formal disapproval never had to accomplish. By the time of withdrawal in December 2018, the proposal had spent five years as announced intention without legal effect. Advisory committees can expose a disagreement in a single public hearing and force a decision within a defined review clock. The preemption asymmetry exposed a disagreement that no committee could settle, that the courts settled only by allocating loss, that the agency attempted and failed to unsettle by rulemaking, and that Congress left untouched.

Closing Assessment

The delivered system of FDA approval law is a layered construction, and each layer answers a failure of the layer beneath it. The statute states a standard and falls silent on method. Regulation supplies the method, in investigational rules, in the definition of an adequate and well controlled investigation, and in the supplement categories that govern labeling change. Fee legislation supplies the capacity and the clock, renewed on a five year cycle that forces the agency, the industry, and Congress to renegotiate the bargain in public view. Guidance supplies the working detail that reviewers and sponsors actually plan around, binding no one in law and constraining everyone in practice. Post-marketing law supplies the correction mechanism, because the controlled record that wins approval cannot describe the uncontrolled use that follows it. An account confined to any single layer misdescribes the whole.

Two products of that layering deserve the emphasis the brief assigned them. The user fee structure made the regulated industry the majority funder of its own premarket regulator, a result no participant designed in one step and every participant has renewed six times because the alternatives on offer were worse for that participant’s purposes. The preemption asymmetry made the identical molecule compensable or noncompensable according to which application type the dispensed product traveled under, a result no legislature voted on and no court announced as policy. Both outcomes emerged from interaction rather than intention, which is the series thesis this article was built to carry: in implementation law, the operative rule is often the one produced where two instruments meet, not the one written in either instrument alone.

Advisory committees mark the limit of what transparency can settle. A committee can force the evidence into the open, compel the agency to frame its questions in advance, and create the record on which courts and Congress later draw, and a committee cannot decide. The withdrawal cases show both halves of that sentence: a unanimous committee and a Commissioner reaching the same conclusion across eleven months of contested process, and a twelve year interval that no hearing could compress until Congress rewrote the procedure in 2022. The approval system is therefore best read as a machine for converting evidence into decisions under resource constraints that Congress sets and resets, with liability law trailing behind the machine and redistributing its errors according to rules the machine never mentions.

Readers preparing for health law coursework or examination will find this material easier to retain as a sequence of instruments than as a list of doctrines. Working notes stay organized in the free VaultBook legislation study notebook, where the pathway stages, the fee cycle, and the case line can be kept in one place, and recall of the same material can be tested with US government and civics study practice on ReportMedic. The statute rewards the reader who learns it the way the sections above present it: who acts, under which instrument, on what clock, and with what obligation left standing when the action ends.

Frequently Asked Questions

Q: How does FDA approval actually work?

A sponsor first assembles a preclinical record of pharmacology, toxicology, and manufacturing information, then files an investigational new drug application, which takes effect thirty days after receipt unless the agency imposes a clinical hold. Testing proceeds through phases that establish safety and dosing, then effectiveness under controlled conditions. The sponsor next files a new drug application, or a biologics license application for a biological product, and a multidisciplinary review team tests the record against the statutory standard of safety and substantial evidence of effectiveness. The agency acts through an approval letter or a complete response letter identifying deficiencies. Approval then creates continuing duties, including adverse event reporting, required post-marketing studies for some products, and labeling that tracks new risk information.

Q: What are the phases before FDA approval?

Phase I introduces the drug into a small number of human subjects to study safety, metabolism, and dose tolerance. Phase II runs controlled studies in patients who have the target condition, testing preliminary effectiveness and refining the dose. Phase III conducts the large, adequate and well controlled investigations that ordinarily supply the core evidence of effectiveness and broaden the safety record across a wider population. Phase IV is not a phase before approval at all; the name describes post-marketing studies conducted after approval. The phase structure itself comes from the investigational regulations at 21 C.F.R. Part 312 rather than from the text of section 505, which speaks in broader terms of investigations and substantial evidence.

Q: Who pays for FDA approval reviews?

Sponsors pay user fees under the Prescription Drug User Fee Act of 1992, and Congress pays the remainder through appropriations. The Congressional Research Service recorded that fees covered 7 percent of program costs in fiscal 1993 and 75 percent of program total costs in fiscal 2023, with appropriations supplying the balance in each year. Since the 2017 reauthorization the fee has two main parts: an application fee paid on submission, supplying roughly a fifth of revenue, and an annual program fee on approved products, supplying roughly four fifths. The fees purchase review capacity and a published set of performance goals. They do not purchase approval, because the goals promise action on an application rather than a favorable outcome.

Q: What is accelerated FDA approval?

Accelerated approval allows the agency to approve a drug for a serious condition on the basis of an effect on a surrogate endpoint, or on an intermediate clinical endpoint, that is reasonably likely to predict clinical benefit. The agency created the program by regulation in 1992, and Congress codified it at 21 U.S.C. 356(c) in 2012. The sponsor must conduct adequate and well controlled confirmatory trials with due diligence to verify the predicted benefit, and the labeling must disclose that approval rests on a surrogate measure. Accelerated approval is a full approval subject to conditions, not a provisional license with an expiry date. When confirmatory trials fail to verify benefit, the agency may withdraw the approval through the procedures the statute and regulations supply.

Q: Can a drug lose FDA approval after confirmatory trials fail?

Yes. Withdrawal is the enforcement mechanism built into accelerated approval. Under 21 C.F.R. 314.530 for drugs and 21 C.F.R. 601.43 for biologics, the agency may withdraw approval when a required study fails to verify clinical benefit, when other evidence shows the product is not safe or effective under its conditions of use, or when the sponsor fails to conduct a required study with due diligence. The bevacizumab breast cancer indication was withdrawn in November 2011 after confirmatory trials failed to verify benefit, and hydroxyprogesterone caproate was withdrawn in April 2023, twelve years after its approval. Contested withdrawal was slow for decades because sponsors could demand a formal hearing. Section 3210 of the Food and Drug Omnibus Reform Act, enacted in December 2022, codified expedited withdrawal procedures built on notice, explanation, a sponsor meeting, and public comment.

Q: Are FDA approval advisory committee votes binding?

No. An advisory committee advises, and the agency decides. No committee vote approves a drug, refuses an application, or withdraws an indication, and the Commissioner and the responsible center officials retain the statutory authority to act. The agency follows committee recommendations in a substantial share of cases, since referral itself signals genuine scientific uncertainty and a lopsided vote from qualified outside experts carries institutional weight. Departure remains lawful whenever the deciding official explains a different reading of the same evidence. In the bevacizumab proceeding, the Oncologic Drugs Advisory Committee voted 6 to 0 in favor of withdrawal in June 2011, and the legal act of withdrawal was Commissioner Margaret Hamburg’s decision of November 18, 2011, rather than the vote that preceded it.

Q: Can you sue a generic maker over a label after FDA approval?

Ordinarily, a failure to warn claim against a generic manufacturer does not survive. In PLIVA, Inc. v. Mensing, 564 U.S. 604 (2011), the Supreme Court held such claims preempted, because federal law requires the generic label to be the same as the brand label and a generic manufacturer cannot use the changes being effected procedure to strengthen a warning on its own authority. A comparable claim against a brand manufacturer survived in Wyeth v. Levine, 555 U.S. 555 (2009), because the brand manufacturer held that unilateral route. Mutual Pharmaceutical Co. v. Bartlett, 570 U.S. 472 (2013), extended preemption to design defect claims against generic manufacturers. Claims that a generic label actually departed from the reference label present a different question, because sameness is a federal duty as well as a federal shield.

Q: What is right to try compared to FDA approval?

Right to try is a route around the approval system rather than a form of approval. The Right to Try Act, Public Law 115-176, signed on May 30, 2018, permits an eligible patient with a life threatening condition, who has exhausted approved treatment options and cannot participate in a clinical trial, to seek an investigational drug directly from its sponsor after the drug has completed Phase I testing. No agency authorization of the individual use is required, and the sponsor carries no duty to supply the product. The agency expanded access program covers similar ground inside the regulated system, with agency authorization, institutional review board oversight, and the possibility of a clinical hold. Neither route approves the product, and neither generates the evidence that an approval requires.

Q: What is an investigational new drug application?

An investigational new drug application is the instrument, authorized by 21 U.S.C. 355(i) and governed by 21 C.F.R. Part 312, that exempts an unapproved drug from the marketing prohibition so the drug can be shipped and administered in human testing. The sponsor submits the preclinical pharmacology and toxicology record, the proposed protocols, information about the investigators, and commitments covering informed consent and institutional review board oversight. The investigation ordinarily becomes effective thirty days after the agency receives the application, without any affirmative grant, unless the agency imposes a clinical hold. A clinical hold, governed by 21 C.F.R. 312.42, is an order to delay or suspend testing, most often because the record shows an unreasonable and significant risk to subjects or lacks the information needed to assess that risk.

Q: What is a complete response letter from the FDA?

A complete response letter is the action letter, governed by 21 C.F.R. 314.110, that the agency issues when review of an application is complete and the application cannot be approved in its present form. The letter describes every specific deficiency the review identified, notes amendments the agency has not yet reviewed where relevant, and, when possible, recommends actions the applicant might take to place the application in condition for approval. The letter is not a final denial. The applicant must resubmit the application addressing all deficiencies, withdraw the application, or request an opportunity for a hearing on whether grounds exist to deny approval. If the applicant takes none of those steps within one year, the agency may treat the failure to act as a request to withdraw the application.

Q: How long does FDA review of a new drug application take?

Under the user fee performance goals, the agency aims to act on 90 percent of standard original applications within ten months of the filing date and on 90 percent of priority applications within six months. Filing itself occurs sixty days after the agency receives the application, once reviewers decide the submission is sufficiently complete to permit substantive review, so elapsed time from submission to action ordinarily runs roughly two months longer than the headline goal. The goals are commitments recorded in letters transmitted to Congress rather than deadlines written into the approval statute. Missing a goal date does not approve an application by default, and an action within the goal can take the form of an approval letter or a complete response letter.

Q: What is a biologics license application?

A biologics license application is the marketing application for a biological product, filed under section 351(a) of the Public Health Service Act, codified at 42 U.S.C. 262(a). The license issues on a demonstration that the product is safe, pure, and potent and that the facility in which the product is manufactured meets standards designed to assure that the product continues to be safe, pure, and potent. Potency performs the work that effectiveness performs for drugs. Because biological products are made by living systems, the application places distinctive weight on process validation, product characterization, and facility inspection. The Center for Biologics Evaluation and Research reviews vaccines, blood products, and cellular and gene therapies, while the drug center reviews many therapeutic proteins, including monoclonal antibodies.

Q: What is the difference between fast track and breakthrough therapy designation?

Fast track designation, at 21 U.S.C. 356(b), is available for a product intended for a serious condition that shows potential to address an unmet medical need, and the showing can rest on preclinical data, early clinical data, or both. Its benefits are more frequent communication with the agency and rolling review of completed application sections. Breakthrough therapy designation, at 21 U.S.C. 356(a), demands more at the door: preliminary clinical evidence indicating that the product may demonstrate substantial improvement over available therapy on one or more clinically significant endpoints. In return, breakthrough therapy adds intensive guidance on efficient development and the involvement of senior managers and experienced review staff. A product can hold both designations at once, and neither designation changes the statutory standard of substantial evidence that approval requires.

Q: What is priority review at the FDA?

Priority review is an agency designation that shortens the review goal for a marketing application from ten months to six months, measured from the filing date. The agency grants the designation when the product, if approved, would provide a significant improvement in the safety or effectiveness of treating, diagnosing, or preventing a serious condition. The designation grew out of the two tier review policy the agency adopted in 1992 and operates through the user fee performance goals rather than through the approval statute. Priority review changes only the clock. The designation does not lower the evidentiary standard, does not shorten the sponsor trials that precede submission, and does not promise approval. Priority review vouchers, created by statutes beginning in 2007, allow a qualifying sponsor to transfer the shortened clock to a different application.

Q: What happens after a drug is approved by the FDA?

Approval opens a second regulatory life for the product. The sponsor must report serious and unexpected adverse drug experiences to the agency within fifteen calendar days under 21 C.F.R. 314.80, must file annual reports under 21 C.F.R. 314.81, and must complete any post-marketing studies or clinical trials required under 21 U.S.C. 355(o)(3) or imposed as conditions of an accelerated approval. Some products carry a risk evaluation and mitigation strategy under 21 U.S.C. 355-1, with elements such as prescriber certification or restricted dispensing settings. The sponsor must keep labeling aligned with new risk information, using the changes being effected supplement where that route applies. The agency in turn retains authority to require safety labeling changes and, on statutory grounds after due notice and opportunity for hearing, to withdraw the approval.

Q: What is a risk evaluation and mitigation strategy?

A risk evaluation and mitigation strategy, known by the initials REMS, is a required risk management program authorized by 21 U.S.C. 355-1 and created by the Food and Drug Administration Amendments Act of 2007. The agency requires a strategy when it determines that one is necessary to ensure that the benefits of a drug outweigh its risks, a determination that can occur before approval or after approval on new safety information. A strategy can include a Medication Guide or patient package insert, a communication plan directed at health care providers, and elements to assure safe use, such as prescriber certification, pharmacy certification, dispensing only in specified settings, documentation of laboratory results, patient monitoring, or enrollment of each patient in a registry. The sponsor must operate an implementation system and submit assessments of the strategy on the schedule the statute sets.

Q: What does substantial evidence of effectiveness mean?

Substantial evidence is the statutory standard at 21 U.S.C. 355(d). The term means evidence consisting of adequate and well controlled investigations, including clinical investigations, conducted by experts qualified by scientific training and experience, on the basis of which those experts could fairly and responsibly conclude that the drug will have the effect it purports or is represented to have under the conditions of use in its labeling. The regulation at 21 C.F.R. 314.126 describes the design features that make an investigation adequate and well controlled, including a clear statement of objectives, a design permitting a valid comparison with a control, and measures intended to minimize bias. The standard governs the quality of the proof rather than a fixed count of trials, and agency practice can accept one adequate and well controlled investigation together with confirmatory evidence in appropriate circumstances.

Q: Can the FDA withdraw approval of a drug?

Yes, on statutory grounds and through a formal proceeding. Under 21 U.S.C. 355(e), the agency may withdraw approval after due notice and opportunity for hearing when the evidence shows that the drug is unsafe for use under the conditions in its labeling, that the drug is not shown to be safe, that substantial evidence of effectiveness is lacking, or that the application contained an untrue statement of material fact. A sponsor may demand a hearing, and the product ordinarily remains marketed while the contest runs, which is why contested withdrawals have consumed years of agency attention. For accelerated approvals, the regulations at 21 C.F.R. 314.530 and 21 C.F.R. 601.43 add failure to verify clinical benefit and failure to conduct a required study with due diligence as grounds, and the omnibus reform enacted in December 2022 codified expedited withdrawal procedures in the statute.

Q: What is a 505(b)(2) application?

A 505(b)(2) application is a new drug application that relies, for part of its safety and effectiveness case, on investigations the applicant did not conduct and does not hold a right to reference, most often published literature or the agency prior findings for an approved drug. The route suits a product that differs from an approved drug in a way that defeats the sameness requirements of the abbreviated pathway, such as a changed dosage form, a new strength, a modified formulation, or a new use, while sharing enough with the approved product that repeating the entire evidentiary program would waste resources. The applicant must supply the bridge, scientific evidence connecting the relied upon findings to the product actually proposed. Because the application is a full new drug application in form, it passes through the same filing decision, review clock, and pair of action letters as an application built entirely on original studies.

Q: Why does Congress renew FDA user fees every five years?

The fee authority was written with a sunset from the beginning. The Prescription Drug User Fee Act of 1992 authorized collections for an initial five year period, and each reauthorization, in 1997, 2002, 2007, 2012, 2017, and 2022, reset the revenue levels, revised the fee structure, and adopted a fresh set of performance commitments negotiated between the agency and the regulated industry. The sunset disciplines every participant. The agency must justify the review workforce the fees employ. The industry must decide what the predictable review system is worth to it. Congress gains a recurring legislative vehicle for drug policy that must pass on a deadline, which is why major policy riders travel on fee bills. Because fees covered 75 percent of program costs in fiscal 2023, allowing the authority to lapse would dismantle much of the review enterprise the fees built.