From a biological idea to a medicine, step by step
Bringing a medicine from a biological hypothesis to a prescription is among the most
complex coordinated undertakings in science, spanning a decade or more and drawing on the
judgment of two dozen specialist disciplines. This page sets out that pathway as 12 discrete
steps across 5 stages. Open any step for the work it involves, the decision that closes it,
and what each function contributes along the way.
10–15 yrsTypical time from discovery to approval
~1 in 10Candidates entering Phase 1 that reach approval
23Distinct functions involved along the way
5,000–10,000Compounds screened per approved medicine
Orders of magnitude, drawn from FDA and NIH descriptions of the process.
Real timelines vary enormously by modality, indication and regulatory pathway — an accelerated-approval
oncology programme and a chronic-disease primary-care drug look almost nothing alike.
Primary purpose. Establish that a specific biological molecule or pathway is causally involved in a disease, and that modulating it is likely to produce clinical benefit without unacceptable harm.
What this step involves
Research for a new drug begins in the laboratory. The first question is not “what molecule should we make?” but “what should we aim at?” — a receptor, an enzyme, a transporter, a gene product, or a pathway whose behaviour differs in disease.
Validation is the harder half. A target is validated when there is convergent evidence that engaging it changes the disease: human genetics linking the gene to the phenotype, tissue expression data, animal models in which the target is knocked out or inhibited, and ideally some human evidence from an existing drug or natural experiment.
This is also where the first commercial and regulatory questions appear. Is the population large enough and well enough defined? Is there an approved therapy that any new entrant would have to beat? Is there a plausible endpoint a regulator would accept?
Key activities
Literature, genetic association and omics analysis to nominate candidate targets
Target tractability assessment — is this druggable by a small molecule, an antibody, or another modality?
In-vitro and in-vivo validation, including knockdown, knockout and tool-compound studies
Competitive and intellectual-property landscape review
Early target safety assessment — what is known about on-target toxicity in humans and animals?
Who does what 8 functions
Every function engaged at this step, and what it is accountable for here. Function names link to the full glossary further down the page.
Leads target nomination and validation, designing and running the disease models, knockdown and knockout experiments and pharmacology studies that establish a causal rather than merely correlative link between the target and the disease.
Begins biomarker thinking at the outset: what measurable signal would demonstrate the target is engaged in a human being, and could a responsive patient subgroup be identified prospectively rather than discovered after the fact?
Performs the target safety assessment, reviewing the biology for on-target liabilities — what else the target does, where else it is expressed, and what is observed in humans who lack it through genetic variation.
Assesses tractability: whether the target presents a binding site a small molecule, antibody or other modality could realistically exploit, and what precedent exists across the wider target class.
Characterises the disease itself — incidence, prevalence, natural history and current standard of care — defining the population a programme would serve and the outcomes that carry weight within it.
Sizes the addressable population and maps the competitive and pipeline landscape, framing whether a successful programme would justify the decade of investment it would require.
Sets the research data-integrity expectations that make early work defensible years later — laboratory notebook practice, raw data retention, and documentation of how each assay was run and qualified.
Builds the initial programme plan, convenes the functions needed to assess the target, and brings the opportunity to a portfolio decision with a defined budget, timeline and set of success criteria.
Deliverables
Target validation package
Target safety assessment
Initial target product profile (TPP) — the label the programme is trying to earn
Decision gate
Target endorsed for screening, with an agreed target product profile.
Primary purpose. Convert a validated target into a specific molecule with the potency, selectivity, and drug-like properties needed to be given to a human being.
What this step involves
Screening produces hits — molecules that do something to the target. High-throughput screening of compound libraries, fragment-based approaches, DNA-encoded libraries, structure-based design, or for biologics, immunisation and phage display all serve here.
Hits become leads through iterative chemistry. Each cycle makes analogues, tests them against potency, selectivity, metabolic stability, permeability, solubility and early safety counter-screens, and feeds the result back into design. This is the structure–activity relationship loop, and it typically runs for hundreds of compounds.
In parallel the assay infrastructure is built: the bioanalytical methods that will later measure drug concentration in blood, and the formulation science that will make the molecule deliverable.
Key activities
High-throughput, fragment, virtual or phenotypic screening to identify hits
Hit-to-lead triage and lead series selection
Iterative medicinal chemistry against potency, selectivity and ADME properties
In-vitro ADME: metabolic stability, permeability, protein binding, transporter and CYP interaction
Early safety counter-screens, including hERG and cytotoxicity panels
Pharmacokinetic and efficacy studies in animal models
Who does what 8 functions
Every function engaged at this step, and what it is accountable for here. Function names link to the full glossary further down the page.
Designs and synthesises each generation of compounds and owns the structure–activity relationship, balancing potency, selectivity, metabolic stability and physicochemical properties across the evolving lead series.
Runs the potency, selectivity and cellular efficacy assays and the animal pharmacology demonstrating that the molecule produces its effect through the intended mechanism rather than an off-target one.
Characterises absorption, distribution, metabolism and excretion in vitro and in animals, identifies the metabolising enzymes and transporters involved, and builds the prediction of human pharmacokinetics.
Runs early safety screens to deselect liabilities while they remain cheap to abandon — cardiac ion channel, genotoxicity and cytotoxicity signals in particular — and flags structural alerts within the series.
Develops and qualifies the assays that quantify compound and metabolites in plasma and tissue, first to support animal studies and later, once fully validated, to measure clinical samples.
Begins route scouting, salt and polymorph selection and early formulation feasibility, assessing whether the chemistry can be scaled economically and the molecule made stable enough to dose reliably.
Qualifies the external laboratories and CROs generating screening, ADME and early safety data, and holds the standards for method records and traceability so results can be reconstructed and relied upon later.
Manages the design–make–test cycle time, usually the rate-limiting step of discovery, and holds the criteria a compound must meet before it can be considered for nomination.
Deliverables
Lead series with established structure–activity relationship
In-vitro ADME and early safety profile
Proof of mechanism in an animal model
Validated bioanalytical method
Decision gate
Lead series endorsed; compounds progress to candidate selection.
Primary purpose. Choose the single molecule that the organisation will take into formal development, and commit the substantial expenditure that decision triggers.
What this step involves
Candidate nomination is a governance event, not a laboratory one. A cross-functional committee reviews the full data package on the leading molecules and selects one — usually with a backup — to enter development.
The decision weighs the pharmacology against everything that will determine whether the molecule can actually become a product: can it be manufactured at scale and at acceptable cost, is the predicted human dose plausible, is the safety margin adequate, is the intellectual property defensible, and does the predicted profile meet the target product profile?
From this point the cost profile of the programme changes sharply. IND-enabling toxicology and GMP manufacture are expensive, so nomination is the last cheap moment to say no.
Key activities
Assembly of the full candidate data package across pharmacology, ADME, safety and chemistry
Human dose and exposure prediction from animal data
Developability assessment — synthesis, scale-up, stability and formulation
Intellectual property position and freedom-to-operate review
Formal governance review and nomination decision
Who does what 10 functions
Every function engaged at this step, and what it is accountable for here. Function names link to the full glossary further down the page.
Presents the candidate and its backup with the full rationale for selection, including where the series’ remaining liabilities lie and what the backup molecule is intended to mitigate.
Delivers the human pharmacokinetic prediction — clearance, half-life, expected exposure and drug-interaction potential — that underpins the proposed clinical dose and dosing interval.
Sets out the safety profile emerging from the screening work and designs the GLP toxicology programme, including species selection, dose levels and the study durations the intended trial will require.
Assesses developability: whether the molecule can be synthesised reproducibly, formulated stably and manufactured at commercial scale within an acceptable cost of goods.
Advises which of the planned studies must be conducted under GLP, audits the data package underpinning the nomination decision, and begins qualifying the vendors that will perform regulated work.
Drafts the clinical development strategy the candidate would support — the indication sequence, the trial designs, and the evidence any intended label claim would ultimately demand.
Builds the early exposure–response reasoning and the first-in-human dose rationale, including the estimated pharmacologically active dose and the escalation increments proposed around it.
Advises on the likely regulatory pathway, the designations available, and the evidence authorities in each target region will expect — including where precedent exists and where the programme would be breaking new ground.
Runs the governance process and the investment decision, presenting the integrated cross-functional case and the resourcing required to carry the candidate through to first-in-human.
Confirms the commercial case against the current and projected competitive landscape, including what will have changed in the market by the time this product could realistically launch.
Deliverables
Nominated development candidate (and backup)
Human dose prediction
Agreed IND-enabling study plan
Updated target product profile
Decision gate
Candidate nominated and funded into preclinical development.
Primary purpose. Generate the nonclinical safety, pharmacology and manufacturing evidence required to justify giving the molecule to a human being for the first time.
What this step involves
Preclinical research answers basic questions about safety before a drug is given to people. Unlike discovery work, these studies are conducted under Good Laboratory Practice — a formal quality system governing conduct, documentation and archiving, because regulators will rely on them.
The core is repeat-dose toxicology in two species, normally one rodent and one non-rodent, at durations that support the intended clinical exposure. These establish the no-observed-adverse-effect level, which drives the human starting dose, and identify which organs to monitor in the clinic.
Around them sit safety pharmacology of the cardiovascular, respiratory and central nervous systems; genotoxicity; and, depending on the programme and its timing, reproductive toxicology and carcinogenicity. ICH M3(R2) sets out which studies are needed before which stage of clinical development — the package is staged, not completed all at once.
In parallel, manufacturing moves to GMP. The clinical trial material must be made under controlled conditions, with analytical methods, specifications and stability data sufficient to show that what goes into the first patient is what the sponsor says it is.
Key activities
GLP repeat-dose toxicology in a rodent and a non-rodent species
Safety pharmacology core battery — cardiovascular, respiratory, central nervous system (ICH S7A)
Genotoxicity battery (ICH S2(R1))
Toxicokinetics to link exposure to findings
Reproductive and developmental toxicology as required by stage and population
GMP manufacture of drug substance and drug product for clinical use
Analytical method validation, specification setting and stability studies
Determination of the no-observed-adverse-effect level and human equivalent starting dose
Who does what 10 functions
Every function engaged at this step, and what it is accountable for here. Function names link to the full glossary further down the page.
Owns the GLP toxicology programme — repeat-dose studies in two species, the genotoxicity battery and safety pharmacology — and the integrated nonclinical safety assessment supporting first-in-human dosing.
Delivers toxicokinetics and metabolite profiling, establishing the exposure margins between the doses that produced findings in animals and the exposure proposed for the first human subjects.
Scales the synthesis, develops the clinical formulation, validates the analytical methods, sets the specifications and manufactures the clinical supply under GMP.
Validates the assays used for toxicokinetic sample analysis and prepares the validated clinical assay that will measure drug concentration in the first human study.
Converts nonclinical exposure into a proposed human starting dose and escalation scheme, applying the no-observed-adverse-effect level and the safety factors prescribed by regulatory guidance.
Audits the pivotal GLP toxicology studies and the GMP manufacture of clinical supply, issues the study-level compliance statements the application depends on, and verifies that the data will withstand inspection.
Shapes the package against regional expectations and prepares the pre-IND interaction, where the agency’s view on the proposed programme can be obtained before anything is formally filed.
Authors the Investigator’s Brochure and the nonclinical written and tabulated summaries — the documents through which every investigator and every reviewer will come to understand the molecule.
Designs the first-in-human protocol concept against the emerging safety profile, including the population, the intensity of monitoring and the stopping rules the nonclinical findings demand.
Sequences the studies against the clinical timeline; toxicology duration is frequently the critical path to the clinic, and the order in which studies start determines the filing date.
Deliverables
GLP toxicology study reports
Safety pharmacology and genotoxicity packages
Investigator’s Brochure
GMP clinical trial material with certificate of analysis
Proposed first-in-human starting dose and escalation plan
Decision gate
Nonclinical package adequate to support first administration in humans.
Primary purpose. Obtain regulatory permission to administer an investigational product to humans, by demonstrating that the proposed study is reasonably safe and scientifically sound.
What this step involves
In the United States the sponsor submits an Investigational New Drug application. It must contain animal study and toxicity data, manufacturing information, the clinical protocol, data from any prior human research, and information about the investigators who will run the study.
The FDA review team then has 30 days to review the original IND submission. If the agency does not object within that window, the study may proceed. If it has serious concerns it issues a clinical hold, which the FDA describes as rare.
Other regions use a different instrument for the same purpose — a Clinical Trial Application in the European Union under the Clinical Trials Regulation, or equivalents elsewhere — but the substance of the package is broadly similar.
An IND is not a one-time event. It stays open for the life of the programme: new protocols, protocol amendments, annual reports and expedited safety reports are all filed to it.
Key activities
Compilation of the IND or CTA across nonclinical, quality and clinical modules
Pre-IND meeting with the agency to align on the package and the development plan
Protocol finalisation and Investigator’s Brochure issue
Institutional review board or ethics committee submission
Clinical trial registration
Who does what 10 functions
Every function engaged at this step, and what it is accountable for here. Function names link to the full glossary further down the page.
Owns the submission: compiles, formats and files the application, manages the pre-IND meeting and all agency correspondence, and tracks the 30-day review clock that governs when dosing may begin.
Authors the protocol, the Investigator’s Brochure and the nonclinical summaries making up much of the package, to the document standards the agency expects of a regulatory submission.
Provides the nonclinical study reports and the integrated safety conclusion stating why the proposed human dose is justified by what was observed in animals.
Provides the chemistry, manufacturing and controls section describing the drug substance, the drug product, the manufacturing process, the specifications and the stability data supporting clinical supply.
Owns the clinical protocol and its medical rationale: the population, the monitoring plan, the risk mitigation, and the medical justification for exposing people to the drug at all.
Justifies the starting dose, the escalation scheme and the stopping rules, and specifies the pharmacokinetic sampling that will answer the questions the study has been designed to answer.
Specifies the design, the sample size and the analysis approach within the protocol, including how dose escalation decisions will be made from data as it accumulates.
Establishes the safety reporting infrastructure before first dose: case intake, the causality assessment process, the expedited reporting obligations and the safety database itself.
Prepares site selection and qualification, essential documents, contracts and the study start-up activities that must all be complete before the first participant can be screened.
Confirms GxP readiness across the sponsor, the manufacturing site and the investigator sites, and verifies that the quality systems described in the application are the ones genuinely in operation.
Deliverables
IND or Clinical Trial Application
Approved protocol and Investigator’s Brochure
Ethics committee or IRB approval
Trial registration record
Decision gate
IND in effect after 30 days without clinical hold; study may begin.
Primary purpose. Determine how the drug behaves in the human body and at what doses it can be given safely — the first and most carefully controlled human exposure.
What this step involves
Phase 1 studies enrol 20 to 100 participants, usually healthy volunteers, though in oncology and some other fields they enrol patients with the disease because the toxicity profile makes healthy-volunteer exposure inappropriate. The FDA notes these studies typically run several months, and that roughly 70% of drugs move to the next phase.
The design is almost always dose escalation. A small cohort receives the lowest dose, is observed intensively, and only then is a higher dose given to the next cohort. Stopping rules and dose-limiting toxicity definitions are specified in advance.
The questions are pharmacokinetic and tolerability, not efficacy: what concentration does a given dose produce, how is the drug cleared, what is the half-life, is exposure proportional to dose, what happens with food, and which adverse events appear as exposure rises. Where possible a pharmacodynamic marker shows whether the target is actually being engaged.
Dedicated clinical pharmacology studies cluster here and continue throughout development — drug-drug interaction, hepatic and renal impairment, QT assessment, and bioavailability work.
Key activities
Single and multiple ascending dose cohorts with predefined escalation rules
Intensive pharmacokinetic sampling and exposure characterisation
Pharmacodynamic and target-engagement biomarker assessment
Food effect, drug-drug interaction and special-population studies
Cardiac repolarisation (QT) assessment
Safety review committee meetings between cohorts
Who does what 13 functions
Every function engaged at this step, and what it is accountable for here. Function names link to the full glossary further down the page.
The lead scientific function in Phase 1. Owns pharmacokinetic and pharmacodynamic characterisation, dose escalation decisions, modelling and simulation, and the dose rationale carried forward into Phase 2.
Provides medical oversight of every participant, reviews the emerging safety data between cohorts, and chairs or supports the dose escalation committee that authorises each increment.
Collects and assesses every adverse event, performs causality assessment against a profile that barely exists yet, and meets expedited reporting obligations for serious unexpected suspected adverse reactions.
Runs the unit or site, manages dosing logistics and the intensive sampling schedules Phase 1 demands, and oversees the vendor network supporting the study.
Analyses the pharmacokinetic samples on which the entire phase depends, under validated methods and with a turnaround fast enough to inform the dose given to the next cohort.
Deploys biomarkers to establish whether the drug reaches its target in humans and engages it — the first direct evidence that the mechanism translates from animals to people.
Builds and cleans the database and manages the rapid data flow required between cohorts, where escalation decisions depend on data being clean within days rather than months.
Files new protocols and amendments to the IND, reports safety information to the agency within required timelines, and handles questions arising during conduct.
Supplies clinical material and manages the formulation changes that commonly occur between cohorts as the dose range, route and tolerability become clearer.
Audits the Phase 1 unit, the bioanalytical laboratory and the sponsor's own GCP processes; establishes the trial master file, deviation handling and vendor qualification that govern the first human exposure.
Primary purpose. Establish whether the drug actually works in the target disease, and at which dose — the first real test of the hypothesis the programme was built on.
What this step involves
Phase 2 enrols up to several hundred people who have the disease or condition, and the FDA notes these studies run from several months to two years. This is the most dangerous phase statistically: only about 33% of drugs move to the next phase.
The design shifts from open-label escalation to controlled comparison. Randomisation, blinding and a control arm — placebo or active — appear here, because efficacy claims require them. Studies are often split into Phase 2a, exploring proof of concept, and Phase 2b, refining dose.
Dose-ranging is the central task. Several doses are usually studied in parallel so that the exposure–response relationship can be characterised rather than guessed, since carrying the wrong dose into Phase 3 is one of the most expensive mistakes in the industry.
Endpoints at this stage are frequently surrogate or intermediate measures chosen for sensitivity and speed. Part of the work is learning which endpoint, population and effect size the registrational programme should be built around.
Key activities
Randomised, controlled, often blinded proof-of-concept studies
Parallel dose-ranging to characterise exposure–response
Endpoint selection and effect size estimation for Phase 3
Population refinement, including biomarker-defined subgroups
Interim analyses and futility assessment
End-of-Phase-2 meeting with the regulator
Who does what 15 functions
Every function engaged at this step, and what it is accountable for here. Function names link to the full glossary further down the page.
Owns the clinical hypothesis, the protocol design, the endpoint choice and the interpretation that determines whether the programme continues, stops, or changes indication entirely.
Designs the study and the dose-ranging analysis, sizes the trial against the effect worth detecting, and runs the interim and futility analyses that may end the programme early.
Models exposure–response across the studied doses and recommends the dose and regimen for Phase 3 — a decision that is expensive and frequently impossible to revisit later.
Delivers enrolment across a larger and more geographically distributed site network, and manages CRO and vendor performance against both timeline and quality metrics.
Tests the biomarker hypotheses and whether a definable subgroup responds better than the overall population, a finding that can reshape the Phase 3 population entirely.
Maintains the developing safety profile, runs signal detection against a growing dataset, updates the Investigator’s Brochure and supports the periodic safety reports.
Manages a substantially larger and more complex dataset through to database lock, including external data arriving from central laboratories, imaging vendors and devices.
Produces the analysis datasets and statistical outputs to CDISC standards, establishing the data structures the registrational programme will inherit and build upon.
Authors protocols, amendments, study reports and the briefing package for the end-of-Phase-2 meeting, which frames how the agency will see the programme going forward.
Prepares and runs the end-of-Phase-2 interaction, which sets the terms of the registrational programme: endpoints, population, comparator and the size of the safety database.
Advances the formulation towards the commercial presentation and generates the comparability data showing that Phase 3 material behaves like the material already studied.
Runs risk-based GCP audits of investigator sites, CROs and central laboratories, oversees trial master file completeness and CAPA management, and begins building the inspection-readiness posture the programme will need at registration.
Begins payer evidence planning and identifies the outcomes, comparators and quality-of-life instruments reimbursement bodies will expect, since omitting them from Phase 3 forecloses the option entirely.
Builds scientific relationships with the investigator and expert community who will enrol, deliver and ultimately interpret the Phase 3 programme, and gathers their view of the proposed design.
Manages the decision point and the transition into a programme that will typically cost an order of magnitude more than everything spent on the molecule so far.
Deliverables
Proof of concept — or a decision to stop
Recommended Phase 3 dose and regimen
Defined target population and primary endpoint
End-of-Phase-2 meeting minutes and regulatory agreement
Decision gate
Efficacy signal, dose and population sufficient to commit to Phase 3.
Primary purpose. Confirm efficacy in a large, representative population and characterise the safety profile well enough to support approval and a label.
What this step involves
Phase 3 studies enrol 300 to 3,000 volunteers who have the disease or condition and run one to four years. The FDA notes that roughly 25–30% of drugs move to the next phase. These are the registrational trials: the evidence on which approval turns.
Scale changes everything operationally. A Phase 3 programme may run across hundreds of sites in dozens of countries, requiring site activation, training, drug supply, monitoring and data flow on an industrial scale, usually with contract research organisations delivering much of it.
The statistical design is fixed in advance and defended afterwards. The primary endpoint, the analysis population, the handling of missing data and the multiplicity strategy are all pre-specified, because a post-hoc result will not support a claim.
Safety becomes a volume problem rather than a case-by-case one. With thousands of exposed patients, the programme can detect less common adverse events, and the pharmacovigilance system must aggregate, code and analyse them continuously rather than only at the end.
Much of the rest of the submission is built in parallel: commercial-scale manufacturing and process validation, the health economic dossier, and the evidence that will support launch.
Key activities
Two adequate and well-controlled confirmatory trials, or one plus confirmatory evidence
Large-scale site activation, training and monitoring across regions
Pre-specified primary, secondary and safety analyses
Independent data monitoring committee oversight
Long-term safety follow-up and open-label extensions
Process validation and commercial-scale manufacture
Patient-reported outcome and health economic data collection
Who does what 17 functions
Every function engaged at this step, and what it is accountable for here. Function names link to the full glossary further down the page.
Owns the medical strategy and the integrity of the registrational programme, and the benefit–risk argument the resulting data must ultimately support in front of a regulator.
Carries the largest operational burden of the programme: hundreds of sites across many countries, CRO oversight, enrolment delivery, drug supply and the timeline itself.
Owns the statistical analysis plan, the pre-specified analyses and the multiplicity strategy — finalised before unblinding and binding thereafter — along with the integrated summaries.
Builds the SDTM and ADaM datasets and the tables, listings and figures for the submission, to standards the agency will load directly into its own review systems.
Runs data cleaning, medical coding and reconciliation across hundreds of thousands of data points, through to the database lock that precedes unblinding.
Manages case processing volume, aggregate safety reporting, signal detection across the blinded dataset, and the risk management plan that will accompany the application.
Authors protocols, study reports and the integrated clinical summaries forming the core of the submission and the basis of the reviewer’s understanding of the programme.
Conducts population pharmacokinetic and exposure–response analyses supporting the dose, the special-population recommendations and much of the dosing section of the eventual label.
Completes process validation at commercial scale, generates the stability data supporting shelf life, and assembles the commercial manufacturing package.
Produces, labels, randomises and distributes clinical supply across every participating country, including comparator sourcing and the import and customs work that entails.
Audits investigator sites, CROs and central laboratories on a risk basis across every participating country, manages deviations and CAPAs at scale, and brings the trial master file and the contributing facilities to the standard a pre-approval inspection will apply.
Prepares scientific communication, publication planning and congress strategy, and maintains the expert relationships that shape how the results will be received.
Primary purpose. Assemble and file the complete evidentiary case that the product is safe and effective for a defined use, manufactured to a reproducible standard.
What this step involves
A New Drug Application is intended to demonstrate that a drug is safe and effective for its intended use. For biologics the instrument is a Biologics License Application; in Europe it is a Marketing Authorisation Application.
The FDA specifies that the application must include preclinical data through Phase 3 trial data — reports on all studies, data and analyses — proposed labelling and directions for use, safety updates, drug abuse information, patent information, data from studies conducted outside the United States, and institutional review board compliance information.
Submissions are organised in the Common Technical Document format: Module 1 regional administrative information, Module 2 summaries, Module 3 quality, Module 4 nonclinical study reports, and Module 5 clinical study reports. It is filed electronically, and a large application runs to hundreds of thousands of pages.
The agency first decides whether the application is complete. If it is not, the team can refuse to file it — an outcome that costs months and is avoided by careful pre-submission work.
Key activities
Pre-submission meeting and agreement on application content
Authoring of Module 2 summaries and the integrated summaries
Compilation and publishing of the electronic Common Technical Document
Drafting of the proposed prescribing information
Risk management plan and, where required, REMS proposal
Submission and filing review by the agency
Who does what 11 functions
Every function engaged at this step, and what it is accountable for here. Function names link to the full glossary further down the page.
Owns the submission end to end: strategy, content plan, publishing, filing and all agency correspondence, including the format and structure the application is required to follow.
Authors the clinical overview and clinical summaries making up Module 2 — the documents a reviewer reads first, and from which much of their impression of the programme forms.
Delivers the integrated efficacy and safety analyses and the datasets reviewers will re-analyse independently, which must reproduce the sponsor’s results exactly.
Produces submission-ready datasets, define files and reviewer guides to CDISC standards, so the agency can run its own analyses without first reconstructing the data.
Owns the benefit–risk argument and the clinical content of the application, including how the programme’s limitations are characterised rather than left for a reviewer to discover.
Provides the integrated safety summary, the risk management plan and the proposed pharmacovigilance activities, including any studies to be conducted after approval.
Provides Module 4 and the nonclinical written and tabulated summaries, including the carcinogenicity and reproductive toxicology work completed during clinical development.
Provides the clinical pharmacology summary supporting dosing, drug interactions, hepatic and renal impairment, and the special-population language proposed for the label.
Confirms data integrity and inspection readiness across every contributing site, laboratory and vendor, and assembles the audit history the agency will ask for as part of the application.
Coordinates a submission involving nearly every function simultaneously against a fixed filing date, where any single late component delays the entire application.
Deliverables
Filed NDA, BLA or MAA in eCTD format
Proposed prescribing information
Risk management plan
Filing acceptance from the agency
Decision gate
Application accepted for review and a review goal date set.
Primary purpose. Allow the regulator to independently assess whether the evidence supports the claimed benefit and an acceptable risk profile, and whether the product can be manufactured reliably.
What this step involves
Once the application is filed, the FDA review team has 6 to 10 months to decide whether to approve the drug. Products granted Priority Review have a goal of agency action within 6 months, reflecting a potential significant improvement in safety or effectiveness.
The review is divided by discipline. Each reviewer examines their own section — the medical officer and statistician review the clinical data, a pharmacologist reviews the animal studies — and each technical discipline also receives a supervisory review. The project manager compiles the individual reviews and the inspection report into an action package; the team recommends an outcome and a senior official makes the decision.
Inspections run in parallel. FDA inspectors visit clinical study sites looking for evidence of fabricated, manipulated or withheld data, and manufacturing facilities are inspected before approval to confirm the process can be run as described.
Where questions need further consideration, the agency may convene an advisory committee for independent expert advice and public comment. These committees include a patient representative. Their recommendations are not binding, but they carry substantial weight.
Throughout, the agency issues information requests on tight timelines. If the review identifies deficiencies that cannot be resolved, the outcome is a Complete Response Letter setting out what would be required.
Key activities
Discipline reviews across clinical, statistical, nonclinical, clinical pharmacology and quality
Mid-cycle and late-cycle communications with the sponsor
Responses to information requests, often within days
Pre-approval inspection of manufacturing sites and bioresearch monitoring of clinical sites
Advisory committee meeting, where convened
Labelling negotiation
Who does what 10 functions
Every function engaged at this step, and what it is accountable for here. Function names link to the full glossary further down the page.
Manages the review: receives and triages agency questions, coordinates responses within days, and negotiates the wording of the label clause by clause.
Answers clinical questions and defends the benefit–risk argument, including in person at an advisory committee where the questioning is public and unscripted.
Responds to statistical questions and conducts re-analyses, frequently within days, as reviewers test the robustness of the primary result against alternative assumptions.
Addresses safety questions, refines the risk management plan, and supports the drafting of the warnings, precautions and contraindications taking shape in the label.
Supports quality questions and hosts the pre-approval inspection of the manufacturing facility, where the process must be demonstrated to operate exactly as filed.
Leads inspection readiness across the clinical and manufacturing sites selected for inspection, manages the conduct of the inspection itself, and owns the response to any findings issued.
Answers dosing, interaction and special-population questions, which commonly reshape the dosing section of the final label more than any other category of review question.
Prepares scientific materials and coordinates expert engagement around the advisory committee, within the strict limits on what may be discussed before an approval exists.
Primary purpose. Convert a regulatory authorisation into an available medicine — manufactured, priced, reimbursed, distributed and supported with accurate information.
What this step involves
If the FDA determines that a drug is safe and effective for its intended use, the agency and the applicant work together to develop and refine the prescribing information — the labelling that describes the basis for approval and how the drug should be used. The label is the legal boundary of everything the company may subsequently claim.
Approval is not availability. Commercial-scale manufacture and distribution must be in place, and in most markets a reimbursement decision stands between authorisation and patient access. Health technology assessment bodies evaluate clinical and economic value on their own timelines, which can extend well past approval.
Launch also changes the company’s obligations. Promotional material becomes subject to regulation, medical information must be able to answer unsolicited enquiries, and the pharmacovigilance system moves from a controlled trial population to the general population.
Where risks require it, approval may carry a Risk Evaluation and Mitigation Strategy or specific post-marketing requirements and commitments, which become conditions of continued marketing.
Key activities
Final labelling and packaging artwork
Commercial manufacture and distribution readiness
Pricing, reimbursement and health technology assessment submissions
Field medical and commercial team training on the approved label
Medical information service readiness
Implementation of REMS or risk minimisation measures
Who does what 10 functions
Every function engaged at this step, and what it is accountable for here. Function names link to the full glossary further down the page.
Finalises the label, secures approvals in further regions on their own independent timelines, and maintains the authorisation through its post-approval commitments.
Scales commercial manufacture, builds the distribution and cold-chain network where required, and positions inventory in each market ahead of the first prescription being written.
Transfers and validates the commercial process at the manufacturing sites and manages the post-approval changes that begin almost immediately after launch.
Leads reimbursement and health technology assessment submissions, which determine real-world access and in many markets matter more to patients than the approval itself.
Deploys field medical teams for scientific exchange with treating clinicians and stands up the medical information service that answers unsolicited questions from healthcare professionals and patients.
Executes the launch plan strictly within the limits of the approved label, where the boundary between promotion and scientific exchange is legally enforced.
Transitions to post-marketing surveillance, including spontaneous report intake from a general population far larger and far less selected than any trial cohort.
Maintains GMP compliance across the commercial manufacturing and distribution network, releases each batch to market, and stands up the complaint handling process that launch will immediately begin to exercise.
Produces the labelling documents, the medical information content and the launch publications that place the registrational data into the peer-reviewed literature.
Coordinates the cross-functional launch sequence, where manufacturing, regulatory, medical and commercial readiness must all converge on a single date.
Primary purpose. Detect what the controlled trials could not, maintain the benefit–risk balance across the life of the product, and extend or defend its use as evidence accumulates.
What this step involves
Clinical trials, however large, study a selected population under controlled conditions. Once a medicine is in general use it reaches patients who are older, sicker, taking other medicines, and more numerous than any trial programme. Rare adverse events, long-latency effects and interactions frequently emerge only here.
Phase 4 studies enrol several thousand volunteers who have the disease or condition and continue to assess safety and efficacy after approval. Some are commitments made to the agency at approval; others are undertaken to support new populations, comparative evidence or reimbursement.
Surveillance runs continuously. MedWatch serves as the gateway for reporting problems with medical products, feeding the adverse event reporting system that supports signal detection. The FDA’s Sentinel Initiative supplements this with active surveillance drawing on large electronic health databases — health records, insurance claims and registries — to monitor approved products in near real time.
Findings have consequences. The agency can require cautions to be added to dosage or usage information, and for more serious issues take stronger measures. Changes to formulation, labelling or dosage strength generally require approval through a supplemental application. Manufacturing sites remain subject to inspection, sometimes unannounced, and a facility that fails to meet standards can be shut down.
Lifecycle management also means growth: new indications, new populations including paediatrics, new formulations and new regions, each of which is its own regulatory submission supported by its own evidence.
Key activities
Spontaneous adverse event collection, coding and expedited reporting
Signal detection, evaluation and benefit–risk assessment
Periodic safety update reports and risk management plan updates
Post-authorisation safety and efficacy studies
Phase 4 and real-world evidence studies
Label updates and supplemental applications
Line extensions, new indications and paediatric programmes
Ongoing GMP inspection and product quality surveillance
Who does what 12 functions
Every function engaged at this step, and what it is accountable for here. Function names link to the full glossary further down the page.
Becomes the central function: owns case intake, causality assessment, signal detection, aggregate reporting, benefit–risk evaluation and risk management across the product’s entire marketed life.
Designs post-authorisation safety studies and real-world evidence programmes, and supplies the denominators and background rates that make a reported signal interpretable at all.
Maintains the authorisation worldwide, files supplements and variations for every change, and negotiates label updates as the safety profile continues to develop.
Handles scientific exchange and medical information in the field, and generates evidence outside the registrational programme where clinical questions remain unanswered.
Leads new indication and line-extension development, which takes the product back through the earlier steps of this pathway with a molecule that is already well characterised.
Supports observational and Phase 4 analyses and the statistical side of signal evaluation, where confounding by indication is the central methodological problem.
Maintains and extends the reimbursement case as comparative evidence accumulates, as competing products enter the market, and as payers revisit decisions they made at launch.
Sustains GMP compliance across a global supply network, manages deviations, complaints and recalls, and maintains the audit and inspection programme covering every vendor touching the marketed product.
Manages commercial supply continuity, manufacturing site transfers and post-approval process changes, all of which must be executed without interrupting patient access or triggering a shortage notification.
No single group carries a molecule from idea to pharmacy shelf. These are the functions that appear across the pathway above, what each one is responsible for, and where in the process it matters most.
Discovery Biology & Pharmacology
Identifies and validates the biological target, builds the disease models, and generates the pharmacology package that argues a molecule should affect the disease.
Studies absorption, distribution, metabolism and excretion — what the body does to the drug. Sets the basis for dose prediction and drug-interaction risk.
Chemistry, Manufacturing and Controls. Owns the synthesis route, formulation, analytical methods, stability and specifications — everything that makes the product reproducible.
Develops and validates the assays that measure drug and biomarker concentrations in biological samples, supporting both nonclinical and clinical studies.
Translates exposure into dose. Owns PK/PD, dose selection and justification, drug-drug interaction and special-population studies, and modelling and simulation.
Owns the clinical strategy and the medical content of the programme: development plan, protocol concepts, endpoints, benefit–risk and interpretation of results.
Collects, assesses and reports adverse events; runs causality assessment, signal detection and benefit–risk; owns safety reporting obligations and risk management.
Maintains the quality systems and audits against GLP, GCP and GMP, and prepares the organisation, its vendors and its sites for regulatory inspection. Quality oversight runs the length of the pathway: it sets data-integrity expectations in discovery, qualifies the laboratories and CROs that generate regulated data, audits toxicology studies and manufacture, conducts risk-based GCP audits of trial sites throughout clinical development, and sustains GMP compliance for commercial supply.
The pathway above is the map. The next question is which function on it fits the training
you already have — and what the first role inside it actually looks like.
Descriptions here reflect the US pathway as the reference case, with EU
equivalents noted where the process differs materially. Figures quoted as ranges are the agencies' own.
Last reviewed October 2026.