External Control Arms in Clinical Trials: Case Studies of Health Authority Submissions

In recent years, the landscape of clinical trials has been evolving, with innovative methodologies gaining traction. One of such advancement is the use of External Control Arms (ECAs) in clinical trials. ECAs utilize existing patient data external to the planned clinical trial as a comparison group.


An ECA can be constructed using data from previous clinical trials, observational studies, or real-world evidence (RWE) such as electronic health records (EHR), registries, or claims data.


An ECA approach is generally used when it is not possible to select a placebo or a standard of care (SoC) group as controls. Using an ECA can expedite the drug development process, reduce costs, and address ethical concerns. This method can be particularly beneficial in rare diseases, oncology, and conditions with unmet medical needs where enrolling a sufficient number of patients for a traditional randomized clinical trial (RCT) may be challenging or unethical.


The use of ECAs in clinical trials has been increasingly recognized by health authorities. Both the U.S. Food and Drug Administration (FDA) and the European Medicines Agency (EMA) have accepted and assessed drug application including an ECA incorporating real-world data (RWD) and historical control data. Nevertheless, due to several limitations emerging from study design a data quality, the ECA component has still a limited impact on the decision-making and approval process.


Common criticism that has been addressed to ECA include:


1. Lack of Randomization: ECAs do not involve random assignment, which can lead to selection bias. This lack of randomization can result in imbalances in baseline characteristics between the experimental and control groups.


2. Data Heterogeneity: Data from ECAs may come from various sources. Differences in data collection methods, patient populations, and treatment protocols can lead to heterogeneity, making comparisons less reliable.


3. Differences in SoC: The SoC may change over time, and historical controls might have received different treatments compared to current patients. This can result in biased comparisons and affect the validity of the trial outcomes.


4. Confounding Variables: ECAs may not account for all confounding variables that could influence patient outcomes. This can obscure the true effect of the investigational treatment.


5. Data Quality and Completeness: The quality and completeness of data used in ECAs can be questionable. Inconsistent or missing data can undermine the reliability of the comparisons.


6. Regulatory Standards: Health authorities may have stringent requirements for the evidence used to support drug approvals. ECAs might not meet these standards, leading to challenges in gaining regulatory approval.


7. Generalizability: Results from studies using ECAs might not be generalizable to the broader patient population due to differences in patient demographics, disease severity, and other factors.


8. Interpretation Challenges: Comparing outcomes between an experimental group and an external control group can be challenging to interpret due to the above factors. This can complicate the assessment of the treatment's true efficacy and safety.


9. Regulatory and Clinical Guidelines: Regulatory and clinical guidelines often prioritize RCTs as the gold standard for evaluating new treatments. External control arms are seen as less robust compared to RCTs.


Despite acknowledged limitations, the number of submissions including ECAs has increased and will continue to grow.


Here are some examples of trials completed for drug approval and including an ECA:


1. Blincyto (Blinatumomab) – Ph-ALL

  • Condition: Philadelphia chromosome-negative relapsed or refractory B-cell precursor acute lymphoblastic leukemia (ALL).
  • Approval: FDA (2014 and 2019 for label expansion), EMA and Health Canada (2015 and 2018 for lebel expansion in EU).
  • Study Design: Single-arm trial of 185 patients, compared against a historical control group of 1139 patients treated between 1990 and 2014.
  • Impact: The ECA was a main component of the clinical evaluation for the efficacy and significantly influenced the regulatory decisions.


2. Blincyto (Blinatumomab) – MRD1

  • Condition: Minimal residual disease (MRD) in ALL.
  • Approval: FDA and Health Canada for both children and adults, EMA for adults only.
  • Study Design: The RWD came from a registry outside the United States.
  • Impact: The ECA had a lower impact on the decision-making process due to uncertainty about the magnitude of the comparative effectiveness.


3. Avelumab (Bavencio)

  • Condition: Metastatic Merkel cell carcinoma (mMCC).
  • Approval: FDA, EMA, and Health Canada (2017).
  • Study Design: Single-arm study of 204 patients, with response rates compared to an ECA of 87 historical patients.
  • Impact: While the ECA faced critiques, the FDA and other agencies still granted approval based on the single-arm study results.


4. Balversa (Erdafitinib)

  • Condition: Locally advanced or metastatic urothelial carcinoma (mUC) with FGFR3 genetic alterations.
  • Approval: FDA (2019), Health Canada (2020).
  • Study Design: Single-arm study compared to an ECA generated from US EHRs and German registry data.
  • Impact: The ECA had limited influence due to methodological concerns, but the FDA granted approval based on safety profile and response rates from the single-arm study.


5. Rozlytrek (Entrectinib)

  • Condition: Adults with non-small cell lung cancer (NSCLC) caused by an abnormal ROS1 gene.
  • Approval: FDA (2019), EMA and Health Canada (2020).
  • Study Design: Integrated analysis of phase II trials and an ECA generated from EHR data.
  • Impact: The ECA had low impact due to perceived methodological issues.


6. Zolgensma (Onasemnogene Abeparvovec)

  • Condition: Spinal Muscular Atrophy (SMA).
  • Approval: EMA (2022).
  • Study Design: Single-arm trial compared against natural history data of untreated SMA type 1 patients.
  • Impact: The ECA provided critical evidence supporting the approval of Zolgensma.


7. Alecensa (Alectinib)

  • Condition: ALK-positive non-small cell lung cancer (NSCLC).
  • Approval: EMA (2017), FDA (2015).
  • Study Design: Real-world data was used to create an ECA of 67 patients to provide additional evidence of efficacy.
  • Impact: The ECA helped achieve full approval in Europe after an initial conditional approval.


8. Trastuzumab (Deruxtecan)

  • Condition: Unresectable or metastatic HER2-positive breast cancer in patients who have failed previous therapy.
  • Approval: FDA (2019), EMA and Health Canada (2021).
  • Study Design: Multiple phase II studies and an ECA built from French hospital EHR data.
  • Impact: The ECA had low impact on regulatory decision-making due to the lack of control for all relevant confounders.


9. Retevmo (Selpercatinib)

  • Condition: RET fusion-positive cancers, including certain thyroid cancers and NSCLC.
  • Approval: FDA (2021).
  • Study Design: The LIBRETTO-001 trial was a single-arm study with results compared to historical control data from patients with similar conditions.
  • Impact: The ECA provided essential comparative data supporting the drug’s efficacy.


10. Idecabtagene (Vicleucel)

  • Condition: Fourth-line relapsed or refractory multiple myeloma.
  • Approval: FDA, EMA, and Health Canada (2021).
  • Study Design: Single-arm study and an ECA generated from RWD across multiple sources.
  • Impact: Despite methodological issues, the ECA helped contextualize the findings supporting approval.




Jahanshahi M, Gregg K, Davis G, Ndu A, Miller V, Vockley J, Ollivier C, Franolic T, Sakai S. The Use of External Controls in FDA Regulatory Decision Making. Ther Innov Regul Sci. 2021 Sep;55(5):1019-1035. doi: 10.1007/s43441-021-00302-y. Epub 2021 May 20. PMID: 34014439; PMCID: PMC8332598.


Jaksa A, Louder A, Maksymiuk C, Vondeling GT, Martin L, Gatto N, Richards E, Yver A, Rosenlund M. A Comparison of Seven Oncology External Control Arm Case Studies: Critiques From Regulatory and Health Technology Assessment Agencies. Value Health. 2022 Dec;25(12):1967-1976. doi: 10.1016/j.jval.2022.05.016. Epub 2022 Jun 25. PMID: 35760714.