New drugs are developed through a long sequence that begins with understanding a disease mechanism and ends, in the minority of cases that succeed, with a product authorized for use and monitored indefinitely thereafter. Between those points lie target identification, compound discovery and optimization, preclinical safety testing, three broad stages of human trials, regulatory review, and the construction of a manufacturing process capable of producing the medicine reliably at scale.
The process typically spans many years, absorbs substantial investment, and fails far more often than it succeeds. Understanding why each stage exists explains both the duration and the attrition.
Stage 1: Target Identification and Validation
Development starts with biology rather than chemistry. Researchers seek a molecular target — most often a protein such as a receptor, enzyme, or ion channel — whose activity contributes to a disease process and which could plausibly be altered by a drug.
Evidence for a target comes from genetics, from studies of disease tissue, from animal models in which the target is deleted or over-expressed, and increasingly from large-scale human genomic data. Target validation asks whether modulating the target actually changes the disease process, and whether doing so is likely to be tolerable, since many proteins serve essential functions elsewhere in the body.
Weak target validation is one of the principal reasons programmes fail later. A compound can be well designed and still have no clinical effect if the underlying target was not driving the disease.
Stage 2: Discovery of a Candidate Molecule
Screening and Hit Identification
Once a target is selected, researchers look for molecules that interact with it. Approaches include high-throughput screening of large compound libraries, fragment-based methods that build up from small binding fragments, structure-based design using the three-dimensional structure of the target, and computational screening. Natural products and existing approved drugs repurposed for new indications are further sources.
Lead Optimization
An initial hit is rarely usable. Medicinal chemists iteratively modify the structure to improve potency and selectivity for the intended target, and to improve the properties that determine whether a molecule can function as a medicine: solubility, stability, absorption, distribution, metabolism, excretion, and the absence of obvious toxicity signals. Selectivity matters because activity at unintended targets is a common source of adverse effects.
Biologics and Other Modalities
Not all medicines are small molecules. Biologics — monoclonal antibodies, therapeutic proteins, and vaccines — are produced in living cells and discovered through different methods, often involving immunization, display technologies, or protein engineering. Newer modalities include nucleic acid therapeutics such as antisense oligonucleotides and small interfering RNA, messenger RNA products, cell therapies, and gene therapies. Each modality has its own manufacturing, delivery, and regulatory considerations, but the overall logic of validation, testing, and review is shared.
Stage 3: Preclinical Development
Before any human receives the candidate, it undergoes laboratory and animal testing designed to establish whether human exposure can be justified.
- Pharmacology characterizes the intended effect and the dose-response relationship in relevant models.
- Toxicology assesses organ toxicity, genotoxicity, and, depending on the intended population, effects on reproduction and development, generally in more than one species.
- Pharmacokinetics describes absorption, distribution, metabolism, and excretion, informing the dose and schedule proposed for humans.
- Formulation and early manufacturing work produces material of defined quality and consistency suitable for clinical use.
Preclinical studies supporting safety are conducted under Good Laboratory Practice. Their results, with manufacturing data and the proposed protocol, form the application to begin human testing: an Investigational New Drug application to the FDA in the United States, or a clinical trial authorisation in European Union member states. An independent ethics committee must also approve the study.
Stage 4: Clinical Development
Human testing proceeds through stages, each answering a different question.
Phase 1
First administration to humans, usually in a small group of healthy volunteers, or patients where the candidate is expected to be toxic. Objectives are safety, tolerability, dose range, and human pharmacokinetics and pharmacodynamics. Doses are escalated cautiously with review between cohorts.
Phase 2
Conducted in patients with the target condition. Seeks a first indication of efficacy, refines dose and schedule, and expands the safety profile in the relevant population. Many programmes end here, either because no effect is seen or because the benefit-harm balance is unfavourable.
Phase 3
Confirmatory trials, usually randomized, controlled, blinded where possible, and multi-centre. Designed with sufficient statistical power to detect clinically meaningful differences in a prespecified primary endpoint and to characterize adverse effects too uncommon to appear earlier. Phase 3 data form the core of the marketing application.
Throughout, trials are conducted under Good Clinical Practice, registered publicly, and monitored for safety, with independent data monitoring committees in larger studies able to recommend early stopping.
Stage 5: Regulatory Review
The sponsor compiles a dossier covering clinical efficacy and safety, non-clinical pharmacology and toxicology, chemistry and manufacturing, and proposed labelling. In the United States this is a New Drug Application or, for biologics, a Biologics License Application. In the European Union it is a Marketing Authorisation Application assessed by the EMA's scientific committees, with the authorisation decision taken by the European Commission.
Regulators evaluate whether the evidence demonstrates a favourable balance of benefits and risks for a defined population and indication, whether the manufacturing process reliably produces a product of consistent quality, and whether the proposed labelling accurately conveys the evidence. They may convene advisory committees, inspect clinical and manufacturing sites, and request further analyses or studies.
Approval is often conditional in practice: restricted to a narrower population than the applicant sought, accompanied by risk-management requirements, or subject to mandated post-authorization studies. Expedited pathways exist for products addressing serious conditions with unmet need, altering timing and evidentiary sequence without removing the requirement to show acceptable benefit-risk.
Stage 6: Manufacturing and Supply
Producing a medicine at commercial scale is a development project in its own right. Processes proven in a laboratory must be transferred to manufacturing scale while maintaining identity, purity, potency, and stability. Facilities operate under Good Manufacturing Practice and are subject to regulatory inspection. Biologics and cell and gene therapies present particular challenges, since they are produced in living systems where small process changes can alter the product.
Supply chain considerations — sourcing of active ingredients, cold chain requirements, packaging, and distribution — are settled before launch, and disruptions in any of them can cause shortages long after approval.
Stage 7: Post-Approval Monitoring
Authorization is not the end of evidence generation. Pharmacovigilance systems collect and assess adverse event reports from clinicians, patients, and manufacturers throughout a product's life. Because pre-approval trials enrol limited and relatively selected populations, rare adverse effects, interactions, and effects in groups under-represented in trials may only become apparent in widespread use.
Phase 4 studies examine long-term safety and effectiveness, wider populations, and comparisons against other treatments. Regulators can respond to emerging evidence by updating labelling, adding warnings, restricting use, requiring additional studies, or withdrawing authorization.
Why Drug Development Takes So Long and Fails So Often
- Biological complexity. Disease mechanisms are incompletely understood, and animal models predict human responses imperfectly.
- Sequential gating. Each stage must complete before the next begins, because exposing larger numbers of people requires the evidence smaller studies provide.
- Efficacy failures. The most common reason for late-stage failure is that the candidate does not produce a clinically meaningful benefit.
- Safety failures. Toxicity may emerge only with longer exposure or in larger, more varied populations.
- Recruitment and duration. Confirmatory trials in chronic diseases can require years of follow-up to observe the outcomes that matter.
- Manufacturing and quality. A product that cannot be made consistently at scale cannot be approved regardless of its clinical performance.
Sources
- U.S. Food and Drug Administration — the drug development process; Investigational New Drug application; New Drug Application
- European Medicines Agency — from laboratory to patient: the journey of a medicine
- U.S. National Institutes of Health — drug discovery and development resources
- International Council for Harmonisation — guidelines on non-clinical safety, GCP, and quality
- World Health Organization — norms and standards for medicines
- ClinicalTrials.gov — clinical study registration