FDA — Drugs, Devices, and the Therapeutic State

The Legal Foundation

The FDA's authority over drugs and biologics flows from a sequence of major statutes:

The Pure Food and Drug Act (1906). The original federal law. Required honest labeling. Did not require safety testing. Did not require efficacy.

The Federal Food, Drug, and Cosmetic Act (1938). Passed in response to Elixir Sulfanilamide. Required pre-market approval based on safety. This is still the foundational FDA statute today, amended many times but never replaced.

The Kefauver-Harris Amendment (1962). Passed in response to thalidomide — a drug that caused thousands of severe birth defects in Europe (it was never approved in the US because FDA reviewer Frances Kelsey demanded more evidence — a story every public health student should know). The amendment required drugs to be proven not just safe but also effective for their intended use, through "adequate and well-controlled investigations."

The Public Health Service Act (1944, amended many times). Provides FDA authority over biologics — drugs derived from living organisms, including vaccines, blood products, gene therapies, and CRISPR drugs. The 1944 Act created the Biologics License Application (BLA) process, parallel to but distinct from the small-molecule New Drug Application (NDA) process.

The Medical Device Amendments (1976). Established FDA authority over medical devices, with the three-class risk-based framework still used today.

The Prescription Drug User Fee Act (1992, "PDUFA"). Allowed FDA to collect fees from pharmaceutical companies to fund drug review. Dramatically accelerated approval timelines and is now the foundation of FDA's operational budget for drug review.

The 21st Century Cures Act (2016). Modernized FDA review processes, accelerated approval pathways, and provided new authorities for regenerative medicine.

This stack of laws is the legal scaffolding that makes the modern FDA possible. The FDA Commissioner's authority isn't derived from the agency's institutional history but from these specific statutory grants, each one written in response to a specific crisis or political moment.

---

The Drug Approval Pipeline

The standard path from molecule to market for a new drug runs through five regulated stages:

Preclinical research. Lab and animal studies establishing basic safety and mechanism. Conducted before any human exposure. Must follow Good Laboratory Practice (GLP) standards. Typically 3–5 years and $10–100 million.

Investigational New Drug (IND) application. Submitted to FDA before any human trials. Contains preclinical data, manufacturing details, and proposed clinical protocols. FDA has 30 days to review; if it doesn't object, trials can proceed. An IND is the gateway to clinical research.

Phase I clinical trials. Small (typically 20–80 healthy volunteers, or for serious diseases like cancer, patients). Primary goal is safety and pharmacokinetics — does the drug do unexpected harm and how does the body process it. Lasts ~1 year.

Phase II clinical trials. Larger (typically 100–300 patients with the target disease). Primary goal is preliminary efficacy and continued safety. Establishes dosing. Lasts 1–2 years. Most drugs fail at this stage.

Phase III clinical trials. Large (typically 1,000–3,000+ patients). Primary goal is statistical proof of efficacy and safety in the target population. Usually randomized and controlled, often double-blinded. Lasts 2–4 years. Costs $100 million to over $1 billion.

New Drug Application (NDA) or Biologics License Application (BLA). Submitted to FDA after successful Phase III. Contains all data from preclinical through Phase III, manufacturing details, labeling proposals, and risk management plans. FDA review typically takes 10 months for standard review or 6 months for priority review.

Phase IV / Post-market surveillance. After approval, the FDA continues to monitor safety. Adverse events are tracked through systems like MedWatch and FAERS. Drugs can be withdrawn from market if post-market data reveals unacceptable risk (Vioxx in 2004, for example).

The total timeline from discovery to approval averages 10–15 years. The total cost — including the cost of all the failed drugs that don't make it through — averages around $2.6 billion per approved drug, according to industry estimates (independent estimates are sometimes much lower, but still well over $1 billion).

This is one reason new drugs are so expensive. It's also one reason regulatory shortcuts — discussed below — are so consequential.

---

Accelerated and Alternative Pathways

The standard pathway is the default but not the only option. FDA has several expedited pathways for drugs addressing serious unmet needs:

Fast Track designation. For drugs treating serious conditions with unmet medical need. Provides more frequent FDA communication and rolling review (submitting sections of the NDA as they're completed rather than all at once).

Breakthrough Therapy designation. For drugs showing substantial improvement over existing therapies in preliminary clinical data. Provides intensive FDA guidance and all Fast Track benefits.

Accelerated Approval pathway. Allows approval based on a surrogate endpoint — a measurable biomarker that's reasonably likely to predict clinical benefit — rather than waiting for the full clinical outcome. Conditional on post-market confirmatory studies. Cancer drugs are frequently approved this way (based on tumor shrinkage rather than waiting for survival data).

Priority Review. Reduces FDA review time from 10 months to 6 months. Granted for drugs offering significant improvements in safety or effectiveness.

Emergency Use Authorization (EUA). Permits use of unapproved products during declared public health emergencies. Used extensively during COVID-19 — every mRNA vaccine, monoclonal antibody, and antiviral was initially under EUA, not full FDA approval. EUA is faster than even accelerated approval but has lower evidentiary standards and is technically not "approval." This will return in BP6.

Compassionate Use / Expanded Access. Allows seriously ill patients to access investigational drugs outside clinical trials. Important for patients with no other options, but limited in scope.

Right to Try (2018). Federal law allowing terminally ill patients to request investigational drugs directly from manufacturers, bypassing FDA. Controversial — many policy scholars view it as more political than substantive, since the existing expanded access pathway already permitted most of what Right to Try authorized.

These pathways exist because the standard 10–15 year approval process is itself a public health cost. Patients with deadly diseases can't always wait. The accelerated pathways are FDA's attempt to balance the safety guarantees of regulation against the urgency of medical need. Whether the balance is right is one of the central debates in FDA policy.

---

Biologics, Devices, and Combination Products

The FDA doesn't just regulate small-molecule drugs. Different product categories have different rules.

Biologics (vaccines, blood products, antibodies, gene therapies, cell therapies) are approved through the Biologics License Application (BLA) process at the FDA's Center for Biologics Evaluation and Research (CBER). Most modern cutting-edge biotech products are biologics. The clinical trial requirements are similar to drugs, but manufacturing is much more complex — biologics are produced by living cells and are extremely sensitive to manufacturing variation. The result is much tighter manufacturing regulation under "current Good Manufacturing Practice" (cGMP) standards.

Biosimilars are FDA-approved versions of biologic drugs that are highly similar to an already-approved reference product. They are not generics in the strict sense — biologics can't be exactly copied due to their complexity. The Biologics Price Competition and Innovation Act (BPCIA) of 2010 created the biosimilar pathway, but uptake has been slower than for traditional generics due to manufacturing complexity and aggressive intellectual property defense by originator companies.

Medical Devices are regulated based on a three-class risk system:

• Class I (lowest risk, like bandages and tongue depressors) — general controls, often exempt from premarket review
• Class II (moderate risk, like blood pressure cuffs and infusion pumps) — usually require 510(k) clearance, demonstrating substantial equivalence to a previously cleared device
• Class III (highest risk, like pacemakers and gene therapy delivery devices) — require full Premarket Approval (PMA) with clinical data, similar to drug approval

The 510(k) pathway is one of the most controversial parts of FDA regulation. It allows new devices to be cleared based on equivalence to existing devices, which can chain back to "predicate devices" that were grandfathered in from the 1970s without rigorous testing. Critics argue this allows insufficient evidence; the FDA argues it appropriately scales review to risk.

Combination products combine drugs, devices, and/or biologics. A gene therapy delivered by a custom device, for example. The FDA's Office of Combination Products determines which center has primary jurisdiction and coordinates review.

Companion diagnostics are tests that determine eligibility for specific drugs (a genetic test that identifies patients who will respond to a particular cancer drug, for example). These are usually reviewed alongside the drug they accompany.

The proliferation of product categories isn't an accident. Modern biotech increasingly produces hybrid products that cross old regulatory lines. The FDA's structure has evolved to handle this, but the lines are continuing to blur.

---

Wait, Actually...

The FDA's reputation depends almost entirely on what it stops, not what it approves.

For the vast majority of drugs the FDA approves, no one outside the patient community ever thinks about FDA review. The reviews happen, the drugs work, and no headlines result. But every time a previously approved drug is withdrawn — Vioxx in 2004 (50,000+ estimated cardiovascular deaths), Avandia restricted in 2010 (cardiovascular risks), Zantac removed in 2020 (carcinogenic contamination) — public confidence takes a hit. Every approval that goes wrong is remembered.

The structural problem: the FDA is rewarded for caution and punished for both excessive caution and insufficient caution. A drug delayed in approval doesn't make headlines unless an activist group makes it one. A drug approved that later causes deaths makes headlines for years. The asymmetric political incentive pushes the FDA toward caution.

That caution costs lives too. Drugs that work but are delayed in approval mean patients who could have benefited didn't. The FDA tries to balance these competing costs, but there is no way to balance them perfectly. Every line drawn between "approve" and "don't approve" — or "approve sooner" and "wait for more data" — produces deaths on both sides of the line.

This is one of the inherent tragedies of regulation. There is no neutral position. Even doing nothing has a cost in lives.

---