Technical writing
FDA Drug Approvals: The NDA, BLA, and ANDA Database Behind Every Drug on the Market
The FDA's Center for Drug Evaluation and Research publishes a complete application-level record of every drug approval action it has taken since 1939—New Drug Applications, Biologics License Applications, and Abbreviated New Drug Applications. That dataset, Drugs@FDA, is the authoritative ledger of the American pharmaceutical market. Every brand-name drug, every generic, every biologic that a patient can legally obtain in the United States traces back to an entry in this database. It is also one of the most structurally rich federal datasets available: it encodes not just approval decisions but the full timeline of supplements, manufacturing changes, labeling revisions, and patent and exclusivity expirations that govern when competitors can enter the market.
What Drugs@FDA contains
CDER publishes Drugs@FDA at accessdata.fda.gov/scripts/cder/daf and distributes bulk data as a set of five flat CSV files: ApplicationsData (one row per application, with sponsor name, application number, and application type), Products (active ingredient, dosage form, route, strength, reference drug flag), Submissions (every action on every application—original approvals, supplements, withdrawals—with action dates and review priority), MarketingStatus (whether a product is currently marketed, discontinued, or temporarily withdrawn), and ApplicationDocs (links to review packages, labels, and approval letters). The same data is queryable in near-real time through the OpenFDA drugs API endpoint at api.fda.gov/drug/drugsfda.json.
Application numbers follow a consistent pattern: NDA022488, BLA125057, ANDA076183. The prefix encodes application type. The number is permanently associated with a specific drug product from a specific sponsor—every supplement, every labeling change, every post-approval study report files under the same number.
The three application pathways
NDA: New Drug Application
An NDA is required for any new small-molecule drug—a new chemical entity (NCE), a new formulation of an existing drug, a new route of administration, or a new combination product. The sponsor must submit a complete clinical package: Phase I safety and pharmacokinetics, Phase II dose-ranging, Phase III randomized controlled trials demonstrating efficacy. The standard review period is 10 months from the date of filing. A Priority Review designation shortens that to 6 months.
NDAs are the most expensive pathway. A full development program for a new molecular entity costs $1–2 billion on average and takes 10–15 years from initial compound identification to first patient prescription. The application itself can run to hundreds of thousands of pages. FDA user fees under PDUFA (Prescription Drug User Fee Act) currently run approximately $4 million per standard NDA.
BLA: Biologics License Application
Biological products—monoclonal antibodies, fusion proteins, vaccines, blood products, gene therapies, cell therapies—require a BLA rather than an NDA. The distinction matters legally: biologics are regulated under the Public Health Service Act rather than the Food, Drug, and Cosmetic Act. Most therapeutic biologics (cancer antibodies, autoimmune treatments, diabetes biologics) are reviewed by CDER. Vaccines, blood products, and tissues fall to CBER—the Center for Biologics Evaluation and Research.
The BLA pathway matters intensely for biosimilar competition. Under the Biologics Price Competition and Innovation Act of 2009 (BPCIA), a reference biologic receives 12 years of data exclusivity from the date of first BLA approval. No biosimilar can be approved on the basis of the reference product's clinical data before that window closes—regardless of when any patents expire. This is substantially longer than the 5-year NCE exclusivity available to small-molecule drugs, and it is the primary reason that biologic drugs remain expensive far longer than their small-molecule counterparts.
ANDA: Abbreviated New Drug Application
Generic drugs enter the market through the ANDA pathway. Because the brand-name drug's safety and efficacy were already established through the NDA, generic applicants do not need to repeat full clinical trials. They must instead demonstrate bioequivalence—that the generic product delivers the same active ingredient to the bloodstream at the same rate and extent as the reference listed drug (RLD). Bioequivalence studies typically enroll 24–36 healthy volunteers in a crossover design and measure pharmacokinetic parameters (Cmax, AUC, Tmax).
Before filing an ANDA, a generic manufacturer must also navigate the patent and exclusivity landscape around the RLD. This is where the Orange Book becomes central.
The Orange Book
The Approved Drug Products with Therapeutic Equivalence Evaluations—universally called the Orange Book after the color of its original print edition—is the FDA's official list of approved prescription and OTC drug products. For every approved NDA and ANDA, the Orange Book records active ingredients, dosage form, route of administration, strength, applicant, and two critical data layers: therapeutic equivalence codes and patent and exclusivity listings.
Therapeutic equivalence (TE) codes tell pharmacists and payers whether a generic can be substituted for the brand without physician intervention. An “A” code (AB, AA, AN, AO, AP, AT) means the product is therapeutically equivalent to the RLD—same active ingredient, dosage form, route, and strength, with demonstrated bioequivalence. A “B” code (BX, BD, BE, BN, BP, BR, BS, BT) means therapeutic equivalence has not been established. Retail pharmacies fill about 90% of brand prescriptions with generic substitutes; TE codes are the legal mechanism that makes that substitution possible.
Patent and exclusivity listings in the Orange Book determine when an ANDA can receive final approval. Brand-name sponsors list patents covering the drug substance, drug product, and any approved method of use. A generic filer must either wait for those patents to expire or file a Paragraph IV certification claiming the patent is invalid or will not be infringed. A Paragraph IV filing triggers a 45-day window in which the brand sponsor can sue for patent infringement, triggering an automatic 30-month stay of ANDA approval while litigation proceeds. The first ANDA applicant to file a successful Paragraph IV certification receives 180 days of generic exclusivity—a period during which no other ANDA for the same product can receive final approval.
Exclusivity types
Exclusivity is separate from patent protection. A patent is a property right granted by the USPTO; exclusivity is a statutory marketing protection granted by FDA. Both can run simultaneously, but they operate independently. The major exclusivity types in Drugs@FDA:
- NCE-1 (5-year new chemical entity): granted on first approval of an active moiety never previously approved in any NDA. No ANDA may be submitted for 4 years; no ANDA may be approved for 5 years. The single most valuable exclusivity type for small-molecule drugs.
- 3-Year (new clinical studies): granted when an NDA supplement contains new clinical data essential for approval—a new indication, new dosage form, new route, or new patient population. Prevents generic approval of that specific change for 3 years but does not block existing generic versions of the drug.
- Pediatric (6-month extension): sponsors who conduct FDA-requested pediatric studies of any drug—brand or generic—receive a 6-month extension of all existing patents and exclusivities. For a blockbuster drug, this extension alone can be worth hundreds of millions of dollars.
- Orphan drug (7-year): drugs treating diseases affecting fewer than 200,000 Americans at the time of designation receive 7-year market exclusivity against approval of the same drug for the same orphan indication. Orphan designation also confers a 50% tax credit on clinical trial costs and waiver of PDUFA fees.
- Biological product exclusivity (12-year): the BPCIA reference product exclusivity described above. Applies to the reference biologic's approved indication; biosimilars to other indications of the same biologic may not be blocked by exclusivity for those other indications.
Application actions and supplements
The Submissions file in Drugs@FDA distinguishes between original approvals and the many subsequent actions that accumulate over a drug's commercial life. An Original Approval (submission_class_code “Type 1” for NCEs, “Type 2” for new active ingredients, “Type 3” for new dosage forms, and so on) is the foundational event. Everything after is a supplement.
Efficacy Supplements (SE) add new indications, new patient populations, or new dosing regimens. Each approved SE typically carries its own 3-year exclusivity and often triggers a new patent filing on the method of use. A drug like a PD-1 checkpoint inhibitor may accumulate dozens of SE approvals over its commercial life, each expanding the addressable market and each potentially resetting portions of its exclusivity clock.
Safety Supplements (SS) add or strengthen warnings, update contraindications, or add Risk Evaluation and Mitigation Strategies (REMS). They may be initiated by the sponsor or required by FDA following post-market surveillance signals. Chemistry, Manufacturing, and Controls (CMC) supplements cover manufacturing site changes, formulation modifications, and scale-up—operationally critical but rarely visible to prescribers.
Priority designations
FDA operates several expedited programs for drugs that address unmet medical needs. These are recorded in Drugs@FDA's submissions data and are heavily referenced in pharmaceutical investor analysis:
- Priority Review: designation granted at time of NDA/BLA submission for drugs offering a significant improvement over available therapy. Reduces the standard 10-month review clock to 6 months. Does not change the evidence standard— only the timeline.
- Breakthrough Therapy: created by FDASIA 2012. More substantive than Priority Review—grants intensive FDA guidance throughout development, including cross-disciplinary project teams and rolling review of completed sections. Intended for drugs showing preliminary clinical evidence of substantial improvement over existing therapy on a clinically significant endpoint.
- Fast Track: the most broadly granted designation. Allows rolling review (FDA reviews completed sections of the NDA as they are submitted rather than waiting for the complete package) for drugs treating serious conditions with unmet need.
- Accelerated Approval: the most controversial pathway. Allows approval based on a surrogate endpoint—a biomarker or other measure that is “reasonably likely to predict” clinical benefit. Post-market confirmatory trials are required to verify actual clinical benefit; FDA can withdraw approval if confirmatory trials fail. Most oncology accelerated approvals have been converted to traditional approval or withdrawn based on confirmatory trial results.
- Orphan Drug Designation: granted pre-approval by FDA's Office of Orphan Products Development. Confers development incentives and, upon approval, the 7-year market exclusivity described above.
Patent cliffs and generic entry timing
The combination of patent expiration dates and exclusivity end dates in the Orange Book defines the “patent cliff”—the moment when a brand-name drug first becomes exposed to generic competition. The economic impact is severe and predictable: within 12 months of first generic entry, the brand-name drug typically loses 80–90% of its unit volume. Generic prices fall to 20–30% of the brand price within the first year as multiple competitors enter, then to 10–15% as the market matures. The brand retains a small share among patients with strong brand preference or insurer formulary structures.
Pharmaceutical companies manage the patent cliff through lifecycle strategies encoded in Drugs@FDA. These include: authorized generics (the brand sponsor launches its own generic on the same day as the first independent generic, splitting the 180-day exclusivity revenue); product switches (introducing a modified-release formulation or fixed-dose combination shortly before patent expiry, then discontinuing the original product); pediatric extensions (requesting FDA pediatric studies to claim the 6-month extension); and new indication filings (SE approvals that carry 3-year exclusivity and new use patents).
The patent cliff calendar drives pharmaceutical company R&D pipeline decisions years in advance. Analysts tracking Drugs@FDA can identify drugs whose exclusivity stacks expire in the next 3–5 years, estimate the resulting revenue loss, and assess whether the sponsor's pipeline contains late-stage replacements. This analysis, built entirely from public regulatory data, forms the foundation of pharmaceutical equity research.
Landmark approvals and controversies
The Drugs@FDA record captures not just routine approvals but several decisions that reshaped the relationship between FDA, industry, and the public.
OxyContin (oxycodone extended-release), NDA020553, approved 1996. Purdue Pharma's original approval relied on a single 12-hour clinical study with no long-term abuse-potential data. The label's claim that the extended-release formulation resulted in “a delay in the absorption of oxycodone” reducing abuse potential was not supported by the evidence submitted. The approval triggered an opioid crisis that killed more than 500,000 Americans over the following two decades. Subsequent NDA supplements added abuse-deterrent formulations and a REMS program; the original label claims were retracted. The DEA ARCOS database documents the downstream distribution of the resulting supply.
Aduhelm (aducanumab), BLA761178, approved June 2021. The most contested FDA approval in decades. Biogen's anti-amyloid monoclonal antibody received accelerated approval based on amyloid plaque reduction as a surrogate endpoint, despite two Phase III trials with conflicting efficacy results and an FDA advisory committee voting 10-0-1 against approval. The initial price was set at $56,000 per year; three members of the advisory committee resigned in protest. CMS subsequently declined to cover Aduhelm for most Medicare patients unless they were enrolled in a clinical trial. The controversy prompted congressional scrutiny of the accelerated approval pathway and resulted in statutory changes under the Consolidated Appropriations Act of 2023, requiring FDA to set timelines for confirmatory trial completion at time of accelerated approval.
Pfizer-BioNTech COVID-19 vaccine (Comirnaty), BLA125742, approved August 2021. The mRNA COVID-19 vaccine operated under Emergency Use Authorization (EUA) from December 2020 before receiving full BLA approval. An EUA is a separate legal authority from BLA approval—it permits use of an unapproved product during a public health emergency when the known and potential benefits outweigh the known and potential risks. The EUA/BLA distinction became significant when federal vaccine mandates for military personnel were challenged in courts on the grounds that EUA products could not be mandated; the BLA approval largely mooted those challenges for the Pfizer product. The FDA does not maintain EUAs in Drugs@FDA's standard bulk data; they appear instead in a separate EUA page on FDA's website.
Wegovy (semaglutide 2.4 mg), NDA215256, approved June 2021 for obesity; Ozempic (semaglutide 1 mg), NDA209637, approved December 2017 for type 2 diabetes. Novo Nordisk's GLP-1 agonist semaglutide illustrates how the same active ingredient can generate multiple NDAs for different indications at different doses. The obesity approval triggered demand that overwhelmed global manufacturing capacity, creating a supply shortage affecting both the obesity and diabetes indications. The supply crisis prompted FDA to place semaglutide on the drug shortage list, temporarily permitting compounding pharmacies to produce the drug—a significant exception to the general prohibition on compounding FDA-approved drugs. FDA removed semaglutide from the shortage list in 2024, triggering legal challenges from compounders. The episode demonstrates how a single NDA action can have cascading regulatory consequences traceable across multiple FDA databases.
Querying the OpenFDA drugs API
The OpenFDA drugs API at api.fda.gov/drug/drugsfda.json exposes the full Drugs@FDA record set with Elasticsearch-style query syntax. Key fields at the top level of each document: application_number, sponsor_name, openfda.brand_name, openfda.generic_name. The products array carries dosage form, strength, route of administration, and marketing status. The submissions array carries the full action history: submission_type (ORIG, SUPPL), submission_number, submission_status (AP for approved, TA for tentatively approved, WD for withdrawn), action_date in YYYYMMDD format, review_priority (PRIORITY, STANDARD), and submission_class_code which encodes the NME classification (“Type 1” through “Type 7” for NDAs, with Type 1 being true NCEs and Type 1-AB covering new biologics).
The following script queries the API for all New Molecular Entity approvals in a given year—filtering to submission_class_code “Type 1” and “Type 1-AB” with an approved status in the target year—and prints a ranked list of brand name, sponsor, application number, and approval date:
import requests
import json
from datetime import datetime
# ---------------------------------------------------------------
# Query the OpenFDA drugs API for New Molecular Entity (NME)
# approvals in a specific calendar year.
# NMEs are submission_class_code "Type 1" and "Type 1-AB" --
# original approvals for new chemical entities or new biologics
# not previously approved in any form.
# ---------------------------------------------------------------
APPROVAL_YEAR = 2023
BASE_URL = "https://api.fda.gov/drug/drugsfda.json"
# OpenFDA date format is YYYYMMDD; action_date lives inside the
# submissions array so we search the nested field directly.
start_date = str(APPROVAL_YEAR) + "0101"
end_date = str(APPROVAL_YEAR) + "1231"
params = {
"search": (
"submissions.submission_class_code:("Type 1" "Type 1-AB")"
" AND submissions.action_date:["
+ start_date + " TO " + end_date + "]"
" AND submissions.submission_status:AP"
),
"limit": 100,
}
resp = requests.get(BASE_URL, params=params, timeout=30)
resp.raise_for_status()
data = resp.json()
results = data.get("results", [])
print(f"Found {len(results)} NME approval records for {APPROVAL_YEAR}\n")
approvals = []
for rec in results:
app_number = rec.get("application_number", "")
sponsor = rec.get("sponsor_name", "")
openfda = rec.get("openfda", {})
brand_names = openfda.get("brand_name", [])
brand = brand_names[0] if brand_names else "(no brand name)"
# Walk submissions to find the NME approval action date
action_date = None
for sub in rec.get("submissions", []):
sc = sub.get("submission_class_code", "")
stat = sub.get("submission_status", "")
ad = sub.get("action_date", "")
if sc in ("Type 1", "Type 1-AB") and stat == "AP" and ad:
# Keep the earliest NME approval in case of multiple subs
if action_date is None or ad < action_date:
action_date = ad
if action_date:
# Parse YYYYMMDD -> readable date
try:
dt = datetime.strptime(action_date, "%Y%m%d")
readable = dt.strftime("%Y-%m-%d")
except ValueError:
readable = action_date
approvals.append({
"brand": brand,
"sponsor": sponsor,
"app_number": app_number,
"date": readable,
})
# Sort by approval date
approvals.sort(key=lambda x: x["date"])
print(f"{'Brand':30s} {'Sponsor':35s} {'App #':12s} Date")
print("-" * 95)
for a in approvals:
print(
a["brand"][:30].ljust(30) + " "
+ a["sponsor"][:35].ljust(35) + " "
+ a["app_number"].ljust(12) + " "
+ a["date"]
)
The API is rate-limited at 240 requests per minute without an API key (1,000 requests per minute with a free key from open.fda.gov). The bulk CSV download from the Drugs@FDA download page is preferable for full-database analysis; the API is better suited for targeted lookups and programmatic monitoring of new approvals as they occur.
Accessing Drugs@FDA data
The Drugs@FDA download page at accessdata.fda.gov/scripts/cder/daf provides bulk downloads as a ZIP archive containing the five CSV files described above. The archive is updated monthly. File sizes are modest—the full ApplicationsData file is under 10 MB—making full local replication straightforward. The Orange Book data is distributed separately through FDA's Orange Book download page as three additional CSV files covering products, patents, and exclusivities.
For researchers building integrated pharmaceutical intelligence, the standard workflow is: ingest Drugs@FDA ApplicationsData and Submissions for approval history; join to Orange Book Patents and Exclusivities on application number to compute patent cliff dates; cross-reference to FDA FAERS adverse events data to track post-approval safety signals; cross-reference to ClinicalTrials.gov for pre-approval trial history and ongoing confirmatory trial status; and cross-reference to FDA 510(k) device clearances where drug-device combination products are involved.
Cross-dataset connections
Drugs@FDA is most powerful in combination with the adjacent FDA databases. The FDA FAERS adverse events database provides post-market safety signals at the individual-report level—the mechanism through which safety supplements and label changes are triggered after approval. FAERS case reports can be linked to NDA/BLA application numbers to track the post-market safety record of specific approved drugs.
ClinicalTrials.gov is the pre-approval counterpart: it catalogs every registered clinical trial, including the Phase II and Phase III studies that appear in NDA/BLA submissions. Trial NCT numbers are sometimes referenced in FDA approval documents; the ClinicalTrials.gov API allows programmatic retrieval of trial design, enrollment, and results data. Sponsors increasingly link their NDA submission summaries to specific NCT numbers in their approval letters and summary basis of approval documents.
FDA 510(k) device clearances cover the medical device analog to drug approvals. Many drug-device combination products—pre-filled syringes, drug-eluting stents, autoinjectors—require both an NDA or BLA for the drug component and a 510(k) clearance or PMA approval for the device component. The 510(k) database is queryable through OpenFDA's device endpoints at api.fda.gov/device/510k.json.
Related writing
FDA FAERS: The Adverse Drug Event Database Behind Post-Market Drug Safety — The FDA Adverse Event Reporting System holds more than 30 million reports of adverse drug reactions. Here is how the seven-file quarterly schema works, how MedDRA terminology organizes reactions, and how disproportionality analysis detects safety signals.
ClinicalTrials.gov: The Federal Registry of 400,000+ Clinical Studies — ClinicalTrials.gov is the pre-approval counterpart to Drugs@FDA—it records the Phase I through Phase III trials that underpin every NDA and BLA submission. Here is the data model, the AACT relational database, and how to query trial enrollment, design, and results programmatically.
FDA 510(k) Device Clearances: The Regulatory Database for 200,000+ Medical Devices — The 510(k) substantial equivalence pathway covers the vast majority of medical devices reaching the U.S. market. Here is how the predicate device chain works, how to query the OpenFDA device API, and how 510(k) clearances intersect with drug-device combination product approvals.
More FDA reference data on this site: the National Drug Code Directory (every marketed drug product), the device classification database (every device type and its risk class), and FDA food and cosmetic recalls.