Clinical Trials Glossary

Definitions are aligned with ICH E6(R2) Good Clinical Practice guidelines, FDA regulations (21 CFR Parts 50, 56, 312), and ClinicalTrials.gov standard terminology.

Trial Phases

Phase 0

Exploratory micro-dose studies in humans, typically involving fewer than 15 participants. Doses are sub-therapeutic — too low to cause side effects or treat disease. The goal is to confirm that a drug behaves in the human body as predicted by preclinical models before investing in larger studies.

Phase 1

First-in-human safety studies. Typically 20–100 participants, usually healthy volunteers (except in oncology, where patients with the target disease are enrolled). The primary goals are to establish a safe dosage range, identify side effects, and understand how the drug is absorbed, distributed, metabolized, and excreted (pharmacokinetics). Approximately 70% of drugs pass Phase 1.

Phase 2

Efficacy and dose-finding studies. Typically 100–300 participants with the target condition. Researchers evaluate whether the drug produces measurable biological effects (pharmacodynamics) at doses that are tolerated, and begin identifying the optimal dose for Phase 3. Approximately 33% of drugs pass Phase 2.

Phase 3

Confirmatory large-scale trials comparing the investigational treatment against the current standard of care or placebo. Typically 1,000–3,000+ participants across multiple sites. Phase 3 data is the primary basis for FDA/EMA regulatory approval. Approximately 25–30% of drugs entering Phase 3 receive approval.

Phase 4 (Post-Market Surveillance)

Studies conducted after a drug receives regulatory approval. Used to monitor long-term safety in the general population, detect rare adverse events that weren't apparent in controlled trials, study use in special populations (elderly, pregnant women, patients with comorbidities), and evaluate effectiveness under real-world conditions.

Trial Design & Methodology

Randomization

The process of assigning participants to treatment groups by chance, eliminating selection bias. Randomization ensures that differences in outcomes between groups are more likely due to the treatment than to pre-existing differences between participants. Methods include simple randomization, stratified randomization (balancing groups by key characteristics), and adaptive randomization.

Blinding (Masking)

A technique to prevent knowledge of treatment assignment from influencing outcomes. Single-blind: only the participant doesn't know their assignment. Double-blind: neither the participant nor the investigator knows. Triple-blind: the data analysis committee is also blinded. Double-blind is the gold standard for reducing bias in subjective outcome measurements.

Placebo

An inactive treatment — such as a sugar pill, saline injection, or sham procedure — that is identical in appearance to the active treatment. Used to measure the placebo effect (improvements due to expectations rather than the treatment itself) and distinguish true drug effects from spontaneous disease improvement. Receiving a placebo does not mean a participant receives "no treatment" — they continue receiving standard medical care.

Crossover Design

Each participant receives both the experimental treatment and the control (usually in different periods, separated by a washout period). This design uses each participant as their own control, reducing variability and allowing smaller sample sizes. Best suited for chronic conditions with stable symptoms.

Adaptive Trial Design

A design that allows pre-specified modifications to the trial or statistical procedures based on accumulating data, without undermining the trial's integrity or validity. Adaptations may include dropping poorly-performing arms, changing sample size, or modifying the randomization ratio. Adaptive designs are increasingly used in oncology and rare diseases to reduce time and patient exposure.

Washout Period

A period before enrollment (or between treatment periods in a crossover trial) during which participants stop taking certain medications. This ensures that prior treatments have cleared the system and won't affect results. The length depends on the half-life of the drugs involved.

Endpoints & Outcomes

Primary Endpoint

The main outcome a trial is designed to measure — the result that determines whether the treatment is considered effective. The trial's statistical analysis is powered to detect a clinically meaningful difference in this endpoint. Examples: overall survival, progression-free survival, HbA1c reduction, NIHSS stroke scale score.

Secondary Endpoint

Additional outcomes measured alongside the primary endpoint, providing supplementary evidence about treatment effects. Secondary endpoints may include quality of life, safety measures, biomarker changes, or other efficacy measures. A trial that meets its secondary but not primary endpoint is generally not considered successful.

Surrogate Endpoint

A measurable biomarker or sign used as a substitute for a clinical outcome that is harder to measure (such as survival). Must be validated as a reliable predictor of the clinical outcome of interest. Examples: tumor shrinkage (surrogate for survival), blood pressure reduction (surrogate for stroke prevention), viral load (surrogate for HIV-related mortality).

Overall Survival (OS)

The time from randomization (or diagnosis) until death from any cause. Considered the gold standard endpoint in oncology trials because it is objective, unambiguous, and directly meaningful to patients.

Progression-Free Survival (PFS)

The length of time during and after treatment that a patient lives without disease progression. Frequently used as a primary endpoint in oncology trials because results are available sooner than OS data. Whether PFS improvements translate to OS improvements is an ongoing area of research.

Hazard Ratio (HR)

A statistical measure comparing the rate of an event (e.g., death, disease progression) between two groups. An HR of 0.70 means the treatment group experiences the event at 70% the rate of the control group — a 30% risk reduction. HR < 1 favors the treatment; HR > 1 favors the control.

Regulatory Designations

IND (Investigational New Drug)

An application submitted to the FDA before clinical testing can begin in the United States. The IND includes preclinical data, manufacturing information, and the clinical protocol. The FDA has 30 days to review an IND before the trial can start; if no objection is raised, the trial may proceed.

Breakthrough Therapy Designation

An FDA designation for drugs that treat a serious condition and show substantial improvement over existing therapies in preliminary clinical evidence. Breakthrough designation provides intensive FDA guidance during development and eligibility for rolling review, significantly accelerating the approval timeline.

Fast Track Designation

An FDA program to facilitate development and expedite review of drugs treating serious conditions with unmet medical needs. Grants more frequent FDA communication and rolling review (submitting completed sections of the NDA before the full application is complete).

Accelerated Approval

An FDA pathway that allows approval based on a surrogate endpoint reasonably likely to predict clinical benefit, rather than direct evidence of clinical benefit. Requires post-approval confirmatory trials to verify the actual clinical benefit. Several cancer drugs have been approved — and later withdrawn — through this pathway.

Orphan Drug Designation

Granted by the FDA to drugs intended to treat rare diseases affecting fewer than 200,000 people in the U.S. Provides 7 years of market exclusivity after approval, tax credits for clinical trial costs, waived FDA fees, and assistance with trial design. A major incentive driving rare disease research.

Safety & Ethics

IRB (Institutional Review Board)

An independent committee — required by U.S. federal law — that reviews and monitors research involving human subjects. The IRB evaluates whether the trial design appropriately protects participant rights, safety, and welfare; whether risks are minimized and reasonable; and whether the informed consent process is adequate. No trial can begin at a U.S. institution without IRB approval.

DSMB / DSMC (Data Safety Monitoring Board)

An independent committee of experts (clinicians, biostatisticians, ethicists) that periodically reviews unblinded accumulating data during an ongoing trial. The DSMB can recommend stopping a trial early if: (1) the treatment shows clear efficacy (early success), (2) significant safety concerns emerge, or (3) the trial is unlikely to reach its endpoint (futility). Required for Phase 3 trials and any trial with significant safety risks.

Informed Consent

A legally and ethically required process through which potential participants receive comprehensive information about a trial — including its purpose, procedures, risks, benefits, alternatives, and their rights — and voluntarily agree to participate. Informed consent is ongoing: participants may withdraw at any time without penalty. The consent document must be IRB-approved and written at an appropriate reading level.

Adverse Event (AE)

Any undesirable medical occurrence in a trial participant, regardless of whether it is considered related to the treatment. Adverse events are graded by severity (Grade 1–5 per CTCAE criteria) and assessed for relationship to the study treatment. All AEs must be systematically recorded and reported.

Serious Adverse Event (SAE)

An adverse event that results in death, is life-threatening, requires hospitalization, causes significant disability, or is a congenital anomaly. SAEs must be reported to the FDA, IRB, and sponsor within strict timeframes (as quickly as 7 days for fatal/life-threatening unexpected SAEs).

GCP (Good Clinical Practice)

An international ethical and scientific quality standard for designing, conducting, recording, and reporting clinical trials. Developed by the International Council for Harmonisation (ICH), GCP ensures that the rights, safety, and well-being of trial participants are protected, and that trial data are credible. Compliance with GCP is required for regulatory submissions in the U.S., EU, and Japan.

Patient & Enrollment

Inclusion Criteria

Characteristics that a person must have to be eligible to participate in a trial. Typically include: specific diagnosis, age range, disease stage or severity, performance status, and adequate organ function. Defined to ensure the study population is appropriate for the research question and that results will be interpretable.

Exclusion Criteria

Characteristics that disqualify a person from participating. Common exclusions include: prior treatment with a similar drug, certain comorbidities (e.g., severe kidney/liver disease), pregnancy or breastfeeding, recent participation in another trial, or certain medications that could interact with the study drug. Exclusion criteria exist to protect participant safety and ensure clean data.

Principal Investigator (PI)

The physician or researcher who leads the trial at a specific site, responsible for the conduct of the study, protection of participants, and integrity of the data collected at that site. Multi-site trials have one overall PI (coordinating investigator) and site-level PIs at each participating institution.

Sponsor

The entity — pharmaceutical company, biotechnology firm, government agency (NIH, DoD), or academic institution — that initiates, manages, and finances the trial. The sponsor is responsible for regulatory submissions (IND), safety monitoring, and data management. Sponsor-investigators are individuals who both initiate and conduct a trial.

CRO (Contract Research Organization)

A company that provides outsourced research services to sponsors, including clinical trial management, data management, biostatistics, regulatory submissions, and site monitoring. CROs handle much of the operational burden of large multi-site trials, allowing sponsors to scale their clinical development capacity without permanent infrastructure.

Biomarkers & Precision Medicine

Biomarker

A measurable biological characteristic — a gene, protein, imaging finding, or other indicator — that can be objectively measured and used as an indicator of a biological state, disease process, or response to treatment. Examples: PSA for prostate cancer, HbA1c for diabetes control, BRCA1/2 mutations for hereditary cancer risk.

Predictive Biomarker

A biomarker that identifies which patients are more likely to respond to a specific treatment. Used to select patient populations for enrichment trials. Examples: HER2 amplification predicts response to trastuzumab; PD-L1 expression predicts response to certain checkpoint inhibitors; EGFR mutation predicts response to osimertinib.

Companion Diagnostic

An FDA-approved in vitro diagnostic test essential for the safe and effective use of a specific drug or biological product. Co-developed and co-approved with the therapy it accompanies. Required before treatment can be prescribed — a patient must test positive on the companion diagnostic to be eligible for the matched therapy.

Basket Trial

A type of platform trial that tests one drug (or drug combination) against multiple diseases or disease subtypes that share a common molecular target or biomarker, regardless of where the cancer originates. Allows efficient evaluation of a biomarker-matched therapy across multiple tumor types simultaneously.

Umbrella Trial

A trial that tests multiple drugs or drug combinations — each matched to a different molecular subtype — within a single disease. Patients are screened for molecular alterations and assigned to the matched treatment arm. Umbrella trials allow efficient biomarker-driven drug development within a single tumor type.

Statistics & Analysis

p-value

The probability of observing a result at least as extreme as the one measured, assuming the null hypothesis (no treatment effect) is true. A p-value of <0.05 is the conventional threshold for statistical significance — meaning there is less than a 5% probability that the observed result occurred by chance. The p-value does not measure the size or clinical importance of an effect.

Confidence Interval (CI)

A range of values, calculated from the sample data, that is likely to contain the true population parameter with a specified level of confidence (typically 95%). A 95% CI for a hazard ratio of 0.70 (0.55–0.90) means that if the study were repeated many times, 95% of the calculated CIs would contain the true hazard ratio. Narrower CIs indicate more precise estimates.

Intent-to-Treat (ITT) Analysis

An analysis strategy in which all randomized participants are included in the analysis according to their original treatment assignment, regardless of whether they completed the treatment, deviated from the protocol, or withdrew. ITT analysis preserves the benefits of randomization and provides a conservative, real-world estimate of treatment effect.

Statistical Power

The probability that a trial will detect a true treatment effect of a specified magnitude, if one exists. Conventional trials are designed with 80–90% power. A trial with insufficient power (underpowered) may fail to detect a real effect — a false-negative result. Power is determined by sample size, effect size, and significance threshold.