2  Lifecycle of a Clinical Trial

The molecule that enters a Phase I trial has already survived years of discovery research and preclinical testing, and the Clinical Study Report that trial produces is just one element of a regulatory submission that synthesizes evidence across multiple studies. In between sits a sequence of interconnected stages, each with its own deliverables and regulatory checkpoints. This chapter maps that sequence from initial concept through regulatory approval, with pointers to the chapters that take each stage in depth.

2.1 The Big Picture

Clinical development, the stretch of the arc in Figure 2.1 where human trials occur, typically spans many years and represents the largest single investment in bringing a drug to market (U.S. Food and Drug Administration 2018; DiMasi, Grabowski, and Hansen 2016).

%%{init: {'flowchart': {'nodeSpacing': 25, 'rankSpacing': 30, 'padding': 8, 'diagramPadding': 20, 'subGraphTitleMargin': {'top': 6, 'bottom': 14}}}}%%
flowchart TB
    subgraph Pre["Discovery and Preclinical"]
        direction LR
        D1["Discovery"] --> D2["Lead<br>Optimization"] --> D3["Preclinical<br>Testing"]
    end
    IND["IND / CTA"]
    subgraph Clinical["Clinical Development"]
        direction LR
        P1["Phase I"] --> P2["Phase II"] --> P3["Phase III"]
    end
    subgraph Post["Regulatory and Post-Market"]
        direction LR
        R1["Submission"] --> R2["Review"] --> R3["Approval"] --> R4["Phase IV"]
    end
    Pre --> IND --> Clinical --> Post

Figure 2.1: Clinical trials within the drug development lifecycle

Most candidates fail during clinical development, which makes the design and execution of each trial consequential for the entire program.

The Pipeline Misnomer

The linear “chevron” model (Figure 2.3) is a useful high-level roadmap and a “deeply engrained misnomer” that connotes a consistent, unidirectional flow (Wagner et al. 2018a). In reality, therapeutic development is a “diverse web of failure-prone iterative learning loops” traversed by scientists, physicians, regulators, payers, and patients.

The Forum on Drug Discovery, Development, and Translation of the US National Academies of Sciences, Engineering, and Medicine proposed the Drug Discovery, Development and Deployment Map (4DM) to capture that complexity (see Figure 2.2). The 4DM replaces the pipeline with a set of interacting “neighborhoods” (the full map has eight; the simplified view here highlights five: basic science, biomarker development, clinical research, regulatory review, and the medical landscape) in which discoveries in one area can render earlier results irrelevant or open new shortcuts to later stages. The map captures this reality with “reiterative loops,” indicating that many “shots on goal” are often required before success is achieved at any step (Wagner et al. 2018a, 2018b).

flowchart LR
    A((Discovery &<br/>Basic Science))
    B((Biomarkers<br/>& Assays))
    C((Clinical<br/>Research))
    D((Regulatory<br/>Review))
    E((Medical<br/>Landscape))

    A <--> B
    B <--> C
    C <--> D
    D <--> E
    E <--> C
    A <--> C
    B <--> D

    style A fill:#e1f5fe,stroke:#01579b
    style B fill:#fff3e0,stroke:#e65100
    style C fill:#e8f5e9,stroke:#1b5e20
    style D fill:#f3e5f5,stroke:#4a148c
    style E fill:#fff9c4,stroke:#f57f17

Figure 2.2: Simplified ‘Neighborhood’ view of the 4DM Map (adapted from Wagner et al., 2018)

Failure rates are not uniform across therapeutic areas. In the Biotechnology Innovation Organization (BIO)’s analysis of 2011-2020 development programs, the overall likelihood of approval (LOA) from Phase I was 5.3% for oncology versus 9.3% for non-oncology programs, with the largest attrition occurring at the Phase II transition (oncology Phase II-to-III success 24.6%) (Biotechnology Innovation Organization, QLS Advisors, and Informa Pharma Intelligence 2021). Oncology carries high scientific uncertainty, heterogeneous patient populations, and intense competition all at once. Academic analyses of trial outcomes reach a similar qualitative conclusion: oncology has among the lowest probabilities of advancing from early clinical testing to approval, even as biomarker-enriched programs can outperform historical averages (Wong, Siah, and Lo 2019).

2.2 Trial Phases at a Glance

The phases are not as sharply bounded as Figure 2.3 and Table 2.1 make them look. Phase I/II combination studies are common, and some programs run Phase II and Phase III concurrently. The conceptual framework remains useful because each phase still has a primary purpose.

%%{init: {'flowchart': {'nodeSpacing': 20, 'rankSpacing': 25, 'padding': 8, 'diagramPadding': 20, 'subGraphTitleMargin': {'top': 6, 'bottom': 14}}}}%%
flowchart LR
    subgraph P1["Phase I"]
        N1["Safety, PK,<br/>tolerability<br/>Is it safe?<br/>~20-100<br/>participants"]
    end
    subgraph P2["Phase II"]
        N2["Efficacy signal,<br/>dose selection<br/>~100-500<br/>participants"]
    end
    subgraph P3["Phase III"]
        N3["Pivotal efficacy<br/>& safety<br/>Confirm<br/>benefit-risk<br/>~300-3,000+<br/>participants"]
    end
    subgraph P4["Phase IV"]
        N4["Long-term safety,<br/>effectiveness<br/>Real-world use<br/>Varies"]
    end
    P1 --> P2 --> P3 --> P4

Figure 2.3: Clinical trial phases and their primary objectives

Table 2.1: Clinical Trial Phases Overview

Phase	Primary Question	Typical Size	Duration	Key Decisions
Phase I	Is it safe in humans?	20-100	6-12 months	Dose escalation, stopping rules
Phase II	Does it show efficacy?	100-500	1-2 years	Go/no-go, dose selection
Phase III	Does benefit outweigh risk?	300-3,000+	2-4 years	Primary endpoint, label claims
Phase IV	How does it perform post-market?	Varies	Ongoing	Safety signals, new indications

The sequence is cumulative. Phase I establishes that the drug can be given to humans at tolerable doses; Phase II produces the first evidence of therapeutic effect and identifies the dose(s) to carry forward; Phase III confirms that the drug works in a larger, more representative population and characterizes the safety profile sufficiently to support labeling. Phase IV continues surveillance after approval. Each phase has a chapter of its own (Chapter 7, Chapter 8, Chapter 9, Chapter 10).

2.3 Anatomy of a Phase III Trial

Phase III trials represent the largest investment and the highest operational complexity, and the lifecycle of a single one (Figure 2.4) is a template that applies, with variations, to earlier phases.

%%{init: {'flowchart': {'nodeSpacing': 25, 'rankSpacing': 35, 'padding': 8, 'diagramPadding': 30, 'subGraphTitleMargin': {'top': 6, 'bottom': 14}}}}%%
flowchart TB
    subgraph Planning["Planning & Design"]
        direction LR
        A1["Protocol<br/>Development"] --> A2["Statistical<br/>Design"] --> A3["Site<br/>Selection"] --> A4["Protocol Authorization<br/>(IND amendment / CTA)"]
    end

    subgraph Startup["Startup"]
        direction LR
        B1["Ethics/IRB<br/>Approval"] --> B2["Site<br/>Contracts"] --> B3["Site<br/>Training"] --> B4["Site<br/>Initiation"]
    end

    subgraph Execution["Execution"]
        direction LR
        C1["Recruitment"] --> C2["Treatment"] --> C3["Data<br/>Collection"]
        C4["Monitoring"]
        C1 -.-> C4
        C2 -.-> C4
        C3 -.-> C4
    end

    subgraph Closeout["Closeout & Reporting"]
        direction LR
        D1["Last Patient<br/>Visit"] --> D2["Database<br/>Lock"] --> D3["Statistical<br/>Analysis"] --> D4["CSR &<br/>Submission"]
    end

    Planning --> Startup --> Execution --> Closeout

Figure 2.4: Lifecycle of a Phase III clinical trial

Planning and Design

A poorly designed protocol creates problems that cannot be fixed during execution, which is why months of work precede the first enrollment. The protocol (the scientific and operational blueprint for the trial) specifies the objectives, design, patient population, treatments, assessments, endpoints, and analysis plan, and drafting it takes clinical, statistical, regulatory, and operational input.

Statistical design (Chapter 11) determines sample size, randomization scheme, and analysis approach, while adaptive and Bayesian methods (Chapter 12) extend it to platform trials and designs that update as data accumulate. Study design (Chapter 13) addresses the broader scientific architecture: parallel vs. crossover, superiority vs. non-inferiority, adaptive vs. fixed. When a concurrent randomized control is impractical, real-world evidence and external control arms (Chapter 14) offer an alternative source of comparison. Protocol benchmarks (Chapter 15) examines what makes a protocol operationally feasible.

Site selection (Chapter 18) identifies the investigators and institutions that will conduct the trial. Regulatory authorization for a Phase III protocol usually rides on an existing IND (as a protocol amendment) in the US, plus Clinical Trial Applications (CTAs) in countries that require them; a new IND is the exception, not the rule, by this stage.

Startup

Startup runs from regulatory approval to first patient enrollment, and it is one of the more variable and time-consuming phases to plan. The work includes obtaining ethics committee/IRB approval (Chapter 4), negotiating site contracts and budgets, training site staff on the protocol and systems, and conducting site initiation visits to verify readiness; the logistics and project management chapter (Chapter 20) covers the operational mechanics of site activation and vendor coordination. Recent Tufts CSDD benchmarking of site activation and patient enrollment gives sponsors reference metrics for setting realistic startup timelines (Lamberti et al. 2024).

Execution

Recruitment (Chapter 19) is frequently the most challenging operational element. Enrollment performance varies widely across studies and sites, and recruitment delays remain a leading cause of timeline slippage (Lamberti et al. 2024). Once patients are enrolled, randomization and blinding (Chapter 16) assign them to treatment arms while protecting the integrity of the comparison, and endpoints (Chapter 17) define what is measured and how success is determined.

Data management (Chapter 21) ensures that the information collected is accurate, complete, and ready for analysis, while monitoring (Chapter 22) provides oversight: verifying that sites are following the protocol, that data matches source documents, and that patient safety is protected. Safety (Chapter 23) is a continuous concern throughout execution, since adverse events must be identified, documented, and reported according to regulatory requirements.

The Clinical Trial IT Infrastructure chapter (Chapter 25) examines the technology systems that carry all of this: Electronic Data Capture (EDC), Clinical Trial Management Systems (CTMS), electronic Trial Master File (eTMF), and Randomization and Trial Supply Management (RTSM). The AI Agents in Clinical Development chapter (Chapter 26) explores how AI is beginning to change operational workflows.

Closeout and Reporting

After the last patient completes the last visit, closeout begins. Database lock freezes the data for analysis. Biostatisticians execute the pre-specified analysis plan. Medical writers prepare the Clinical Study Report (CSR) (Chapter 28), the document that presents the trial’s findings.

The CSR becomes part of the regulatory submission (Chapter 29): the New Drug Application (NDA), Biologics License Application (BLA), or Marketing Authorization Application (MAA) that requests marketing authorization. The essential documents chapter (Chapter 24) covers the documentation requirements that must be maintained throughout and archived after the trial, and the closeout chapter (Chapter 27) addresses the operational side: site closeout visits, document archiving, and the administrative closure of the trial.

2.4 Cross-Cutting Concerns

Four concerns run the entire length of a trial (Figure 2.5).

flowchart LR
    R["Regulatory compliance"]
    E["Ethics and patient protection"]
    Q["Quality and data integrity"]
    C["Economic constraints"]
    L(["Trial lifecycle:<br/>planning through closeout"])
    R --- L
    E --- L
    Q --- L
    C --- L

Figure 2.5: Cross-cutting concerns that span the trial lifecycle

Regulatory compliance shapes everything from initial protocol design to long-term data retention. Ethics and patient protection requires informed consent and IRB oversight maintained from first participant to last. Quality and data integrity rests on adherence to the Attributable, Legible, Contemporaneous, Original, and Accurate (ALCOA) principles. Economic constraints are the multi-billion-dollar investments that define the feasibility of modern clinical development.

2.5 How to Use This Book

Figure 2.6 lays out the structure, and readers with different backgrounds will want different entry points into it. For a conceptual foundation, start with Part I: history (Chapter 1), regulatory (Chapter 3), and ethics (Chapter 4). Part II covers drug development, economics, and the progression through Phases I-IV. Part III treats Phase III design and execution in operational detail, from statistical methods through monitoring and safety, and Part IV covers closeout, the CSR, and regulatory submission. Technology has two chapters of its own: IT infrastructure (Chapter 25) and AI agents (Chapter 26).

%%{init: {'flowchart': {'nodeSpacing': 20, 'rankSpacing': 35, 'padding': 8, 'diagramPadding': 20, 'subGraphTitleMargin': {'top': 6, 'bottom': 14}}}}%%
flowchart TB
    subgraph Part1["Part I: Foundations"]
        F1["History"] --> F2["Lifecycle"]
        F2 --> F3["Regulatory"]
        F3 --> F4["Ethics"]
    end
    
    subgraph Part2["Part II: Phases"]
        P1["Drug Development"] --> P2["Economics"]
        P2 --> P3["Phase I-IV"]
    end
    
    subgraph Part3["Part III: Phase III Deep Dive"]
        direction TB
        D1["Design & Statistics"]
        D2["Recruitment & Logistics"]
        D3["Data & Monitoring"]
        D4["Technology & Innovation"]
    end
    
    subgraph Part4["Part IV: Completion"]
        C1["Closeout"] --> C2["CSR"] --> C3["Submission"]
    end
    
    Part1 --> Part2 --> Part3 --> Part4

Figure 2.6: Book structure and chapter relationships

Clinical trials are both a scientific method and an operational system. The science determines what questions can be answered; the operations determine whether the trial can actually answer them.

Biotechnology Innovation Organization, QLS Advisors, and Informa Pharma Intelligence. 2021. “Clinical Development Success Rates 2011-2020.” https://go.bio.org/rs/490-EHZ-999/images/ClinicalDevelopmentSuccessRates2011_2020.pdf.

DiMasi, Joseph A., Henry G. Grabowski, and Ronald W. Hansen. 2016. “Innovation in the Pharmaceutical Industry: New Estimates of R&D Costs.” Journal of Health Economics 47: 20–33. https://doi.org/10.1016/j.jhealeco.2016.01.012.

Lamberti, Mary Jo, Abigail Dirks, Nicholas Kikuchi, Neha Patel Cervantes, and Kenneth Getz. 2024. “Benchmarking Site Activation and Patient Enrollment.” Therapeutic Innovation & Regulatory Science 58 (4): 696–703. https://doi.org/10.1007/s43441-024-00638-1.

U.S. Food and Drug Administration. 2018. “The Drug Development Process.” https://www.fda.gov/patients/learn-about-drug-and-device-approvals/drug-development-process.

Wagner, John, Andrew M. Dahlem, Lynn D. Hudson, Sharon F. Terry, Russ B. Altman, C. Taylor Gilliland, Christopher DeFeo, and Christopher P. Austin. 2018a. “A Dynamic Map for Learning, Communicating, Navigating and Improving Therapeutic Development.” Nature Reviews Drug Discovery 17 (2): 150. https://doi.org/10.1038/nrd.2017.217.

———. 2018b. “Application of a Dynamic Map for Learning, Communicating, Navigating, and Improving Therapeutic Development.” Clinical and Translational Science 11 (2): 166–74. https://doi.org/10.1111/cts.12531.

Wong, Chi Heem, Kien Wei Siah, and Andrew W. Lo. 2019. “Estimation of Clinical Trial Success Rates and Related Parameters.” Biostatistics 20 (2): 273–86. https://doi.org/10.1093/biostatistics/kxx069.