SEMINAR
Designing Single-Arm Clinical Trials: Principles, Applications, and Methodological Considerations
2025 年 7 巻 3 号 p. 90-98
詳細
2025 年 7 巻 3 号 p. 90-98
Single-arm trials (SATs) are clinical studies without a parallel control group, serving as a vital alternative to randomized controlled trials (RCTs) in scenarios where traditional trial designs are impractical. These trials are particularly relevant in rare diseases, advanced malignancies, novel treatment modalities, and life-threatening conditions, where ethical concerns, logistical challenges, or small patient populations limit the feasibility of RCTs. SATs enable expedited evaluation of therapeutic interventions, often forming the foundation for regulatory approvals.
This article explores the principles, applications, and methodological considerations of SATs. Their advantages include smaller sample size requirements, faster timelines, and regulatory acceptance by agencies such as the U.S. Food and Drug Administration (FDA) and European Medicines Agency (EMA). Despite these benefits, SATs face challenges, such as potential biases due to the lack of a control group, limitations in endpoints, and reliance on historical controls that may compromise result validity. Best practices in SAT design are outlined, including refining scientific questions, defining eligibility criteria, selecting clinically meaningful endpoints, and employing robust statistical methods like Simon’s two-stage design and Bayesian approaches.
Single-arm trials (SATs) lack a control group and are typically open-label, without random-ization or blinding1). Fig. 1 provides a basic diagram illustrating the structure of an SAT. These trials serve as a practical alternative to randomized controlled trials (RCTs) in scenarios where RCTs are unfeasible, such as advanced malignancies, rare diseases, emergent infectious diseases, novel treatment methods, or medical device evaluations2). In oncology, SATs are commonly used during the early stages of drug development, such as Phase I or II trials3,4).
SATs are particularly advantageous in situations where patient populations are small or when immediate treatment is necessary. They are also commonly used as an exploratory step before conducting clinical trials with a control arm, helping researchers refine hypotheses and guide the design of subsequent studies. Below are specific applications of SATs:
1.1.1 SATs in oncologySATs play a significant role in oncology drug development, particularly in countries like China, where they are widely used to expedite the approval of therapies for advanced or relapsed cancers. These trials allow for quicker assessment of treatment efficacy and safety, providing patients with faster access to potentially life-saving medications in urgent clinical scenarios where traditional RCTs may be impractical or time-consuming5).
In addition to serving as standalone evidence for regulatory approval, SATs are often used in early-phase oncology studies (Phase I or II) to explore the feasibility, safety, and preliminary efficacy of new therapies. This exploratory role helps refine dosing regimens, identify patient populations most likely to benefit, and lay the groundwork for later-stage trials with control arms3,4).
1.1.2 SATs in rare diseases and life-threatening conditionsSATs are particularly valuable in the study of rare diseases and life-threatening conditions, where small patient populations or ethical constraints make RCTs challenging. For example, during the COVID-19 pandemic, SATs provided pivotal safety and efficacy data for emergency-use therapeutics, enabling rapid regulatory decisions6). In rare diseases, SATs help establish proof-of-concept and generate data that can inform the design of subsequent controlled trials.
1.1.3 SATs for medical devices and novel therapiesSATs are frequently employed in the evaluation of novel medical devices and treatment methods. They provide quicker insights into safety and efficacy compared to RCTs, enabling faster innovation cycles. Additionally, these trials often serve as an initial step to evaluate feasibility and identify technical or clinical challenges before conducting larger trials with control groups2).
1.2 Advantages Of Single-arm TrialsSATs offer distinct advantages in specific clinical and regulatory contexts, making them an important tool in the design of clinical research. Key benefits include the following:
1.2.1. Ethical considerationsSATs provide a consistent and equitable treatment approach, particularly in life-threatening diseases or rare conditions where no alternative treatment options exist. By bypassing randomization, SATs allow all participants to receive potentially beneficial therapies, aligning with ethical principles that prioritize patient well-being and autonomy2). This is especially critical in scenarios where withholding treatment in a control group would be considered unethical.
1.2.2. Practical advantagesSATs are more feasible in scenarios with limited patient populations or urgent timelines. They typically require smaller sample sizes compared to RCTs, reducing the burden of patient recruitment2). Additionally, SATs can be conducted more efficiently, often achieving research objectives within a shorter timeframe, which lowers costs and expedites the development process1).
1.2.3. Regulatory acceptanceRegulatory authorities, such as the U.S. Food and Drug Administration (FDA) and the European Medicines Agency (EMA), have recognized the role of SATs in providing critical evidence for drug and medical device approvals6). In particular, SATs are often included in regulatory frameworks for conditions where conducting RCTs is impractical. Their use in oncology and rare diseases has led to the approval of innovative therapies based on robust SAT data.
While SATs offer significant advantages, they should be carefully designed and analyzed to ensure reliability and scientific rigor, particularly in contexts where their findings may guide high-stakes clinical and regulatory decisions.
1.3 Limitations Of Single-arm TrialsDespite their advantages, SATs face significant challenges that can impact their validity and generalizability. Key limitations include the following:
1.3.1 Absence of parallel controlsThe absence of a control group introduces potential biases in the interpretation of results. Without randomization or blinding, it is difficult to isolate the true effects of the intervention, as outcomes may be influenced by confounding factors or patient selection2). This lack of scientific rigor often limits the strength of evidence generated by SATs.
1.3.2. Endpoint limitationsSATs often rely on response rates or other surrogate endpoints as primary measures of efficacy. While these endpoints can provide valuable insights, they may not fully capture comprehensive safety data or long-term outcomes such as progression-free survival (PFS) or overall survival (OS)7). This limitation can hinder the ability to assess the intervention’s broader clinical impact.
1.3.3. Limitations of external non-concurrent controlsIn many SATs, historical or external controls are used to provide a point of comparison. However, differences in patient populations, treatment protocols, or data collection methods between the SAT and the historical cohort may introduce bias and complicate result interpretation6).
1.3.4. Ethical concernsAlthough SATs can address ethical issues by avoiding placebo or no-treatment arms, their use in populations with limited treatment options raises concerns about the adequacy of evidence for regulatory approval8). Approving interventions based on SATs alone may lead to uncertainties in the long-term safety and efficacy of the treatment.
1.3.5. Confounding factors and biasSATs are particularly susceptible to confounding factors, such as the influence of patient demographics or disease severity on outcomes. While careful study design and statistical adjustments can help mitigate these biases, achieving robust and reliable results is challenging9).
Despite these limitations, SATs remain a valuable tool in clinical research, particularly when traditional trial designs are impractical or unethical. Recent advances in statistical methodologies, such as Bayesian approaches and sensitivity analyses, have improved the reliability of SAT findings9,10). Regulatory agencies have also established standardized guidelines for SAT design, enhancing their quality and acceptability11,12).
Ruan J et al. conducted an open-label, single-group, uncontrolled study to investigate the efficacy and safety of lenalidomide combined with rituximab in newly diagnosed mantle cell lymphoma (MCL) patients13).
MCL is an incurable disease typically treated with high-intensity immunochemotherapy, such as intensified chemotherapy and autologous stem cell transplantation (ASCT). However, the median age of MCL onset is 65 years, leaving many patients unsuitable for such intensive therapies. This study addressed the unmet clinical need for alternative treatment strategies in this population.
The rationale for combining lenalidomide and rituximab was supported by high-quality literature highlighting their therapeutic potential. The researchers hypothesized that biologic agents might provide effective disease control with a favorable side-effect profile compared to conventional chemotherapy, making them suitable for a broader range of patients.
The study’s primary endpoint was the objective response rate (ORR), with secondary endpoints including PFS, OS, and quality-of-life (QoL) scores. Detailed inclusion and exclusion criteria, treatment regimens, supportive care measures, efficacy evaluation methods, and adverse event recording procedures were outlined in the protocol. A total of 38 patients were enrolled, with the sample size calculated using Simon’s two-stage design.
After a median follow-up of 30 months (range: 10–42 months), the results showed:
• Intention-to-treat (ITT) Population: ORR of 87% and a complete response (CR) rate of 61%.
• Evaluable Population (36 patients): ORR of 92% and a CR rate of 64%.
• Secondary Endpoints: 2-year PFS of 85% (95% Confidence Interval (CI): 67–94) and 2-year OS of 97% (95% CI: 79–99).
Since the SAT lacked parallel controls, the investigators used a historical cohort of outpatient MCL patients receiving chemotherapy as a comparator. They emphasized that the study population closely resembled real-world clinical practice, enhancing the generalizability of the findings. The study concluded that lenalidomide combined with rituximab is effective for newly diagnosed MCL and warrants further investigation.
2.2 Example 2Shitara et al. conducted a series of studies investigating Trastuzumab Deruxtecan (DS-8201a) in patients with HER2-positive or low-expressing advanced gastric cancer. This case focuses on the phase I dose-expansion cohort of DS-8201a for HER2-positive advanced gastric cancer14).
Gastric cancer is the third leading cause of cancer-related deaths globally, with significantly higher incidence rates in East Asia compared to Europe and North America. This geographical difference influenced the study’s target population. Currently, trastuzumab, a monoclonal antibody targeting HER2, in combination with chemotherapy, is the standard of care for HER2-positive advanced gastric cancer. However, newer anti-HER2 agents such as lapatinib, pertuzumab, and trastuzumab emtansine have failed to demonstrate superior efficacy. Through literature analysis, the researchers highlighted two key challenges: the heterogeneity of HER2-positive gastric cancer and the observed downregulation of HER2 protein expression in gastric cancer cells following trastuzumab treatment. These insights formed the basis for investigating DS-8201a, addressing unmet clinical needs in patients whose disease progresses after trastuzumab-based therapy.
The phase I study, conducted after determining the recommended dose during the dose-escalation phase, aimed to expand the patient population to evaluate safety and preliminary efficacy. Despite being exploratory, the researchers calculated a sample size of 40 patients based on historical data (p0 = 10%) and the anticipated objective response rate (p1 = 25%).
Results after a median follow-up of 5.5 months included:
• Safety: All patients experienced at least one adverse event, with grade ≥3 events predominantly involving hematologic toxicity. Severe treatment-related adverse events occurred in 25% of patients, but no drug-related deaths were reported.
• Efficacy: Among the 44 patients enrolled, 19 (43.2%; 95% CI: 28.3–59.0) achieved an objective response.
The authors concluded that DS-8201a demonstrated manageable safety and promising preliminary efficacy in heavily pretreated HER2-positive advanced gastric cancer patients. These findings support further investigation in this population.
Designing an SAT requires careful consideration of multiple elements to ensure scientific rigor, ethical compliance, and meaningful outcomes. The following subsections outline the critical steps involved in planning and executing a robust SAT.
3.1 Define And Refine The Scientific QuestionThe foundation of a well-designed SAT lies in identifying unmet clinical needs, which serve as the rationale for the study. Clearly defining the scientific question ensures the study’s relevance and provides a focused objective. As demonstrated in Examples 2.1 and 2.2, researchers highlighted critical gaps in existing treatment options and emphasized the necessity of breakthrough approaches. These unmet needs were articulated in the background sections of their studies, guiding the development of novel therapeutic interventions.
3.2 Rationale For The Investigational DrugA robust rationale for the investigational drug is essential to ensure the scientific validity and ethical integrity of an SAT. Establishing this rationale involves demonstrating the drug’s potential efficacy and safety based on preclinical data, prior clinical studies, or a strong theoretical foundation. This step not only protects participants but also minimizes the risk of wasting resources on ineffective interventions15). The rationale is particularly critical in combination therapies, where interactions between agents must be justified and supported by evidence. For instance, understanding the synergistic effects, pharmacological mechanisms, and potential toxicity profiles can guide the development of a well-founded hypothesis.
Key considerations when developing the rationale include:
i. Scientific Evidence: Leverage preclinical studies, pharmacodynamic data, and existing clinical evidence to support the drug’s potential benefits.
ii. Unmet Clinical Needs: Align the drug’s mechanism of action with gaps in current treatment options, as seen in the examples of lenalidomide and rituximab for MCL and Trastuzumab Deruxtecan for gastric cancer.
iii. Patient-Centric Approach: Emphasize how the drug addresses patient-specific needs, such as reducing toxicity, improving quality of life, or targeting specific disease subtypes.
By articulating a clear and compelling rationale, researchers can build a strong foundation for the trial, ensuring its relevance and increasing the likelihood of generating meaningful results.
3.3 Define The Study PopulationA well-defined study population ensures the trial’s relevance, generalizability, and scientific validity. Careful selection criteria can reduce confounding factors, enhance the interpretability of results, and align the study objectives with the targeted clinical population.
3.3.1 Inclusion criteriaInclusion criteria specify the characteristics required for participation in the trial. These criteria help ensure the population aligns with the research objectives. Common elements include:
i. Disease or Condition: For example, “patients with newly diagnosed diffuse large B-cell lymphoma (DLBCL).”
ii. Disease Stage or Subtype: Clearly define relevant subtypes or stages, such as early or advanced-stage disease.
iii. Age Range: Set appropriate age limits, such as “patients aged 18–75 years.”
iv. Clinical and Laboratory Parameters: Include measurable indicators, such as “absolute neutrophil count >1,500/mm3” or “eGFR ≥60 mL/min/1.73 m2.”
v. Performance Status: Specify functional status, such as “Eastern Cooperative Oncology Group(ECOG) performance status 0–2.”
vi. Treatment History: Define prior therapies allowed or excluded, such as “no prior systemic therapy for DLBCL.”
3.3.2 Exclusion criteriaExclusion criteria identify characteristics that disqualify individuals from participating. These criteria minimize risks and reduce confounding variables. Common elements include:
i. Comorbidities: Exclude participants with conditions like “uncontrolled diabetes” or “active infection.”
ii. Concurrent Medications: Restrict medications that may interfere with the investigational product, such as “current use of immunosuppressants.”
iii. Pregnancy and Lactation: Exclude individuals who are pregnant or breastfeeding to avoid potential risks to the fetus or infant.
iv. Previous Therapy-Related Issues: Disqualify participants with prior adverse reactions to similar therapies.
v. Substance Abuse: Address adherence challenges, such as “history of alcohol or drug abuse within the past year.”
By clearly defining both inclusion and exclusion criteria, researchers can ensure that the study population is scientifically appropriate and ethically justified, improving the overall quality and reliability of the trial.
3.4 Define Clinically Meaningful EndpointsThe selection of clinically meaningful endpoints is essential for the success of SATs. Endpoints should align with the study’s objectives, providing reliable measures of efficacy and safety while addressing the needs of the patient population and regulatory requirements.
SATs are commonly used in Phase II trials to evaluate preliminary efficacy and are occasionally applied in Phase III settings for specific contexts3,16). These trials often focus on efficacy and safety endpoints, particularly in areas such as oncology and rare diseases, where SATs are prevalent for assessing therapies in life-threatening conditions17).
Key endpoints for SATs include:
i. Primary Endpoints: Measures such as ORR and duration of response (DOR) are widely used, especially in oncology, to provide early evidence of treatment effectiveness.
ii. Secondary Endpoints: PFS, OS, and quality of life are often included to offer a more comprehensive evaluation of the intervention’s impact.
In some cases, SATs incorporate non-inferiority and superiority endpoints to assess novel treatment strategies or explore de-escalation therapies. These endpoints can be valuable in proof-of-concept studies3).
To enhance the relevance of SATs, researchers should also consider endpoints specific to the study population. Examples include survival rates in life-threatening conditions and patient-reported outcomes to capture quality-of-life improvements. By selecting appropriate endpoints, SATs can generate meaningful data to guide clinical and regulatory decisions.
3.5 Statistical Analysis MethodsRobust statistical methods are essential for ensuring the validity and reproducibility of SAT findings. Proper sample size calculation and statistical techniques tailored to the trial’s objectives and endpoints are key components of reliable trial design.
3.5.1 Sample size calculationThe sample size for an SAT should be determined based on the trial’s primary endpoint and predefined thresholds, such as expected efficacy or survival rates. Proper calculation minimizes resource use while maintaining the statistical power to detect meaningful effects. Simon’s two-stage design is a widely used method for SATs, particularly in exploratory studies, as it allows interim assessments to determine whether pre-specified efficacy thresholds are met18).
Several approaches for sample size calculation are summarized in Table 1. These include methods based on ORR, median survival, and binomial testing, among others. For example, Simon’s two-stage design optimizes patient accrual by stopping the trial early if the treatment is deemed ineffective, conserving resources while maintaining rigor.
| Method | Underlying Assumptions | Common Use Cases |
|---|---|---|
| Objective Response Rate (ORR) | Historical response rate (p0), target response rate (p1), α and β values | Estimating response rates in cancer treatments (e.g., tumor shrinkage) |
| Median Survival (e.g., PFS) | Median survival follows exponential distribution; λ1(tatget) and λ0(historical) represent failure rates | Assessing PFS or OS |
| Binomial Test (Single Sample) | Target proportion (p1) compared to historical control (p0); binary endpoints | Evaluating binary outcomes, such as success rates or toxicity levels |
| Simon’s Two-Stage Design | Pre-specified p0 (null hypothesis) and p1(alternative hypothesis); iterative optimization | Exploratory studies to minimize unnecessary accrual of patients for ineffective treatments |
| Bayesian Approach | Combines prior knowledge with observed data; flexible assumptions on distributions | Studies with limited historical data or when a probabilistic framework are preferred |
| Equivalence or Non-Inferiority Testing | Compares proportions with a defined margin (Δ); p = pooled proportion. | Evaluating non-inferiority or equivalence of a new treatment compared to historical controls. |
Key variables defined: p0: Historical response rate or survival rate. p1: Expected or target response rate or survival rate under the new treatment. λ0:Historial failure rates,λ1:Target failure rates (λ = ln(2)/Median Survival). Δ: Non-inferiority or equivalence margin. α (alpha): Type I error rate, representing the probability of falsely rejecting the null hypothesis (false positive). β (beta): Type II error rate, representing the probability of failing to reject the null hypothesis when the alternative hypothesis is true (false negative).
The choice of statistical methods must align with the trial’s objectives, primary endpoints, and data types. A summary of commonly used methods and their applications is presented in Table 2. For instance, Kaplan-Meier survival analysis is frequently used to estimate time-to-event data, such as PFS or OS, while Bayesian approaches are ideal for integrating prior knowledge with observed data.
| Statistical Method | Purpose | Common Applications |
|---|---|---|
| Descriptive Statistics | Summarize baseline characteristics, treatment outcomes, and adverse events | Evaluating patient demographics, disease characteristics, and ORR |
| Binomial Test | Assess whether observed proportions differ from a predefined historical control | Testing ORR or binary outcomes (e.g.,success/failure) |
| Kaplan-Meier Survival Analysis | Estimate survival probabilities (e.g., PFS, OS) over time | Analyzing time-to-event data (e.g., PFS or OS) |
| Log-Rank Test | Compare survival curves to a historical benchmark or predefined threshold | Evaluating whether the survival distribution differs significantly from expectations |
| Cox Proportional Hazards Model | Analyze the effect of covariates on survival time | Adjusting for patient characteristics to evaluate treatment effects in time-to-event outcomes |
| Bayesian Analysis | Incorporate prior knowledge and update probabilities with observed data | Flexible analysis of treatment effects when historical data or expert opinions are available |
| T-Test (or Z-Test) | Compare the mean of a continuous variable to a predefined value or historical control | Evaluating outcomes (e.g., tumor size reduction or biomarker changes) |
| Wilcoxon Signed-Rank Test | Test the median of paired data or non-parametric continuous data | Assessing outcomes like changes in tumor size or non-normally distributed continuous variables |
| Confidence Interval Estimation | Provide a range of plausible values for key parameters (e.g., response rates, survival rates) | Assessing precision and reliability of estimates (e.g., ORR, PFS, or OS) |
| Adverse Event Analysis | Summarize and classify treatment-related toxicities | Monitoring safety and tolerability profiles of the intervention |
| Regression Analysis | Explore relationships between outcomes and predictors (e.g., age, staging, biomarkers) | Identifying factors influencing treatment response or survival outcomes |
Each method has its strengths and limitations. Combining multiple approaches can enhance the robustness of SAT analyses. For example, descriptive statistics provide an overview of baseline characteristics and treatment outcomes, while Kaplan-Meier analysis and log-rank tests offer detailed insights into time-to-event data.
By employing these statistical tools effectively, SATs can generate meaningful data to guide clinical and regulatory decision-making, ensuring high-quality evidence despite the absence of a control group.
3.6 Ethical ConsiderationsSATs must adhere to ethical principles, including obtaining informed consent from participants and ensuring that the study design minimizes risks while maximizing potential benefits. Although SATs may have ethical advantages over RCTs, such as avoiding the need for placebo or control arms in life-threatening conditions, these benefits must be balanced against the need for scientific rigor.
Key ethical considerations include:
i. Transparency: Clearly communicate the study’s purpose, risks, and benefits to participants.
ii. Risk-Benefit Analysis: Ensure that the anticipated benefits outweigh the risks for the participant population.
iii. Independent Oversight: Involve ethics committees to monitor study compliance and address emerging concerns.
A strong ethical foundation is critical to preserving the integrity of the study and its results while respecting the rights and welfare of participants.
3.7 Selecting Historical ControlsTo emulate the principle of comparability inherent in RCTs, SATs often rely on historical data as external controls. The selection of historical controls is crucial to the validity of the study, requiring careful consideration of the data’s quality and relevance.
Criteria for Historical Control Data:
i. Time Periods: Data should come from similar timeframes to ensure comparability in treatment approaches and clinical practices.
ii. Disease Backgrounds: The characteristics of the historical cohort must closely resemble those of the SAT population, including disease stage, demographics, and baseline health.
iii. Sample Size: The historical data set should be sufficiently large to provide robust comparisons.
The reliability of historical data must also be critically analyzed1). High-quality sources, such as RCTs, systematic reviews, or meta-analyses, should be prioritized to reduce biases and enhance the study’s credibility.
3.8 Early Collaboration In Trial DesignEngaging key stakeholders early in the study design process is essential to optimize trial methodology and ensure adherence to ethical and scientific standards. Collaboration fosters a multidisciplinary approach, integrating diverse expertise to address complex challenges in SAT design.
Key Stakeholders and Roles:
i. Clinicians: Provide insights into patient populations, clinical endpoints, and treatment feasibility.
ii. Statisticians: Design robust analytical frameworks and ensure appropriate statistical methods are applied.
iii. Ethics Committees: Review the study design to ensure compliance with ethical guidelines and safeguard participant welfare.
iv. Drug Development Teams: Align the trial objectives with regulatory and commercial goals, ensuring that the study produces actionable results.
By fostering collaboration, researchers can enhance the quality, efficiency, and impact of the trial, ultimately generating meaningful evidence to inform clinical practice and regulatory decisions.
SATs offer distinct advantages in evaluating the preliminary efficacy of new therapies, particularly in scenarios where traditional RCTs are impractical. However, their inherent limitations, such as the absence of a control group and potential biases, must be meticulously addressed during the design phase to ensure valid and reliable results.
Key priorities for successful SATs include identifying unmet clinical needs, employing rigorous study designs, and utilizing robust statistical methods tailored to the study’s objectives. These elements not only enhance the scientific validity of SATs but also ensure their relevance to clinical practice and regulatory decision-making.
Looking ahead, the integration of genomics, biomarkers, and real-time imaging technologies holds significant potential to optimize SAT designs. These innovations can provide deeper insights into treatment mechanisms, improve patient stratification, and contribute to the advancement of precision medicine. Future research should continue to refine these approaches, expanding the utility and impact of SATs in addressing complex clinical challenges.
The authors thank Ms. Li Kong, President of Academy of Clinical Research and Study, for coordinating this research work and communication.
No potential competing interests relevant to this article are reported.
Shuna Yao, MS: Conceptualized the study, conducted the literature review, and drafted the initial manuscript.
Qingyao Shang, MD, PhD: Provided clinical expertise, contributed to the design of the study, and critically reviewed the manuscript for clinical accuracy.
Meishuo Ouyang, MD, PhD: Assisted in data analysis, interpreted clinical trial methodologies, and contributed to the manuscript revision.
Heng Zhou, MD, PhD: Contributed to the methodological considerations, provided insights on regulatory approval processes, and reviewed the manuscript for scientific rigor.
Zhihua Yao, MD, PhD: Participated in the study design, conducted background research on single-arm trial applications, and edited the manuscript.
Yanyan Liu, MD, PhD: Assisted in the development of the study rationale, contributed to the discussion on statistical methods, and finalized the manuscript formatting.
Sheng Luo, MD, PhD: Supervised the study, designed the statistical analysis framework, provided biostatistical expertise, and oversaw the final manuscript preparation and submission
Sheng Luo is one of the Editorial Board members of Annals of Clinical Epidemiology. This author was not involved in the peer-review or decision-making process for this paper.
