When a new drug finally appears on pharmacy shelves, the journey behind it typically spans 10 to 15 years and involves thousands of patients, hundreds of researchers, mountains of regulatory filings, and costs that the Tufts Center for the Study of Drug Development has estimated, in aggregate, at over $2.5 billion per approved compound (including the cost of failures). Understanding how that process works helps make sense of medical news — why a "promising" drug in early trials may still be years from approval, why the FDA sometimes moves faster on some diseases than others, and what post-market drug safety surveillance actually does.
The primary authority over prescription drug approval in the United States sits with the Food and Drug Administration's Center for Drug Evaluation and Research (CDER). The process it administers has evolved significantly since the 1962 Kefauver-Harris Amendments, which were passed in the aftermath of the thalidomide tragedy in Europe, where the drug caused severe birth defects in thousands of infants. Those amendments required pharmaceutical companies to demonstrate efficacy — not just safety — before approval, and established the basic clinical trial framework still in use today.
Step One: Discovery and Preclinical Research
Drug development begins long before any human being takes the compound. Researchers — in academic laboratories, government research centers like the National Institutes of Health (NIH), or pharmaceutical and biotechnology companies — identify a biological target (typically a protein, enzyme, or receptor implicated in a disease) and screen thousands or sometimes millions of chemical or biological compounds to find ones that interact with that target in potentially useful ways.
Promising candidates then enter preclinical testing: laboratory studies in cell cultures and animal models designed to assess whether the compound has the hoped-for biological activity, and to establish an initial safety profile. Preclinical studies examine pharmacokinetics (how the body absorbs, distributes, metabolizes, and excretes the compound), pharmacodynamics (what the compound does to the body), and toxicology (at what doses does it cause harm). Most compounds fail at this stage — they prove toxic, ineffective in animal models, or chemically impossible to turn into a stable drug product.
The IND Application: Entering Human Testing
A company that wants to begin human testing must submit an Investigational New Drug (IND) application to the FDA. The IND is not a request for approval; it is a request for authorization to proceed with clinical testing. It must include the preclinical data, a detailed description of the proposed clinical protocols, information about the facilities where trials will be conducted, and evidence of adequate informed consent procedures.
The FDA has 30 days to review the IND and raise any safety objections. If no objections are raised — or once they are resolved — clinical trials can begin. The IND remains active throughout the development program; the sponsor must submit regular safety updates and notify the FDA of any serious adverse events that occur during trials.
Phase 1 Trials: Safety in Healthy Volunteers
Phase 1 trials are primarily about safety, not efficacy. They typically enroll 20 to 100 participants, most often healthy volunteers rather than patients with the target disease (the exception is oncology, where trials typically enroll cancer patients because experimental cancer drugs are too toxic for healthy volunteers). Phase 1 trials answer basic questions: How does the human body handle this compound? What dose range is tolerable? What side effects appear and at what doses? How quickly is the drug absorbed and eliminated?
The classic Phase 1 design is a dose-escalation study: participants receive progressively higher doses, with careful monitoring after each escalation, until the maximum tolerated dose is identified or a predetermined dose ceiling is reached. In the United States, Phase 1 trials are typically conducted at specialized Phase 1 units at academic medical centers or dedicated clinical pharmacology facilities.
Roughly 70% of drugs that enter Phase 1 trials proceed to Phase 2, though that figure varies considerably by therapeutic area.
Phase 2 Trials: Does It Work?
Phase 2 trials are the first time the drug is tested in patients with the target disease, and the first opportunity to gather preliminary evidence of efficacy. They typically enroll hundreds of participants. Phase 2 trials have two primary goals: establishing that the drug produces the biological or clinical effect hoped for, and refining the dosing regimen for subsequent trials.
Phase 2 is often divided into Phase 2a (early, exploratory efficacy studies) and Phase 2b (later, more rigorous dose-finding studies with controlled designs). Many drugs fail at Phase 2 — they may be safe but fail to demonstrate meaningful efficacy, or they may produce unacceptable side effects at effective doses. The approximately 33% success rate from Phase 1 entry to approval reflects how sharply the odds thin at this stage.
"Clinical trials are not a bureaucratic formality — they are the only systematic way to tell the difference between a drug that works and one that merely seems to work in early, uncontrolled observations. History is littered with treatments that looked promising before controlled trials revealed they were useless or harmful." — Paraphrased from writings of physician-researcher Ben Goldacre
Phase 3 Trials: The Pivotal Test
Phase 3 trials are the large, definitive studies that form the evidentiary backbone of an FDA approval application. They typically enroll thousands of patients at multiple sites across multiple countries, and are almost always randomized controlled trials — participants are randomly assigned to receive the new drug or a comparator (either an existing standard-of-care treatment or, where no standard treatment exists, a placebo).
Randomization is critical because it controls for the many factors that could otherwise make a treatment look better or worse than it actually is. Blinding — ensuring that neither participants nor researchers know who is receiving which treatment — controls for placebo effects and observer bias. Together, these design elements make a well-conducted randomized controlled trial the most reliable instrument available for determining whether a treatment actually works.
Phase 3 trials measure clinically meaningful endpoints: survival, disease progression, symptom control, quality of life. They also continue to collect safety data, and for the first time provide information about how the drug performs across diverse subgroups — different ages, sexes, races, and patients with comorbidities.
The NDA: Submitting for Approval
When Phase 3 results are available, a company seeking approval submits a New Drug Application (NDA) — or, for biologics like monoclonal antibodies, a Biologics License Application (BLA) — to the FDA. The NDA is an enormous document: it contains all preclinical data, all clinical trial data, proposed labeling, manufacturing information, and inspection-ready facility data. NDAs frequently run to hundreds of thousands of pages.
The FDA assigns the application a review priority — standard review (10-month target) or priority review (6-month target, for drugs that offer major advances over existing therapy or address unmet medical needs). For many applications, the FDA convenes an independent advisory committee of outside experts to review the data and provide a recommendation, though the agency is not bound by the committee's vote.
FDA reviewers examine the data for substantial evidence of efficacy (typically requiring at least two adequate and well-controlled Phase 3 trials), an acceptable safety profile relative to the drug's benefits, and evidence that the manufacturing process will consistently produce a safe, pure, and potent product.
Expedited Pathways
The FDA has developed several programs to accelerate access to drugs for serious conditions where there is unmet medical need:
- Fast Track Designation facilitates more frequent communication with the FDA during development and allows rolling review — the FDA begins reviewing completed sections of the NDA before the full application is submitted.
- Breakthrough Therapy Designation, created by Congress in 2012, provides all the benefits of Fast Track plus more intensive FDA guidance. It is granted when early clinical evidence indicates the drug may offer substantial improvement over existing therapy.
- Accelerated Approval allows the FDA to approve drugs based on a surrogate endpoint — a biological marker reasonably likely to predict clinical benefit (such as tumor shrinkage as a surrogate for survival) — rather than waiting for the final clinical outcome. Post-market confirmatory trials are required.
- Priority Review Voucher programs incentivize development of drugs for neglected tropical diseases and pediatric rare diseases by awarding transferable vouchers that can be used to accelerate review of another, commercially valuable drug.
Post-Market Surveillance: The Fourth Phase
FDA approval does not end the oversight process. Phase 4, or post-market surveillance, continues for the life of an approved drug. Manufacturers are required to submit periodic safety reports, and the FDA's MedWatch program collects voluntary adverse event reports from healthcare providers and patients. The FDA Adverse Event Reporting System (FAERS) receives hundreds of thousands of reports per year and is analyzed for safety signals.
Phase 3 trials, even large ones, cannot detect rare adverse events that occur in fewer than one in several thousand patients. They also cannot fully characterize the drug's behavior in elderly patients, pregnant women, or people with uncommon comorbidities, who are often underrepresented in trials. Post-market surveillance is the mechanism for identifying these issues after approval.
When post-market data reveals a serious safety problem, the FDA can require label changes, risk evaluation and mitigation strategies (REMS) — restricted distribution or required monitoring programs — or in severe cases can withdraw approval entirely. Rofecoxib (Vioxx), withdrawn in 2004 after post-market data revealed elevated cardiovascular risk, is a frequently cited example of both the limitations of pre-market trials and the functioning of the post-market safety system.
Generic Drugs and the ANDA Pathway
When a brand-name drug's patent protection expires, generic manufacturers can seek approval through an Abbreviated New Drug Application (ANDA). Because a generic must demonstrate that it contains the same active ingredient, in the same dosage form and strength, and is bioequivalent to the brand-name drug — meaning it delivers the same amount of drug to the bloodstream at the same rate — it does not need to repeat the full clinical trial program. The FDA's Office of Generic Drugs reviews ANDAs and has worked to reduce the backlog of applications that built up in the early 2010s.
Bioequivalence is determined primarily through pharmacokinetic studies: volunteers take both the generic and brand-name drug, and blood samples are taken at intervals to compare how the drugs are absorbed. The FDA's standard requires that the 90% confidence interval for the ratio of the generic's and brand-name's key pharmacokinetic parameters falls within 80% to 125% of each other — a range that clinical research has shown produces no meaningful difference in therapeutic effect for most drugs.