When Context Is Hard to Come By: External Comparators and How to Use Them

Abstract

External comparators, also referred to as historical or synthetic controls, present transformational opportunities for broad context and insights alongside clinical research results. The recent confluence of access to quality real-world data (RWD), advanced epidemiologic methods, and legislative directives to regulators for expanded use of RWD is increasing interest in real-world external comparators, opening the door to achieve broader generalizability and learn more, faster. In this less standardized area of research, tailored scientific methodology must be applied for external comparators to accomplish clinical development objectives. Here, we describe methodological considerations for design and illustrate how RWD comparators have been used for regulatory and reimbursement decisions.

Application of Real-World Data to External Control Groups in Oncology Clinical Trial Drug Development

“…As noted by Gray et al., to ensure an internally valid comparison, the external and internal populations should ideally be exchangeable with each other with respect to: eligibility criteria, patient characteristics, mode of treatment (e.g., surgery, chemotherapy, radiation therapy), outcome measure, time period, and setting ( 6 ). However, as external data are typically collected using different methods and from different sites than where the trial is being conducted ( 36 ), such criteria are unlikely to be met. This is particularly true when using RWD, which are accrued without the same level of uniformity as clinical trial data and are not based on a common protocol ( 36 ).…”

“…However, as external data are typically collected using different methods and from different sites than where the trial is being conducted ( 36 ), such criteria are unlikely to be met. This is particularly true when using RWD, which are accrued without the same level of uniformity as clinical trial data and are not based on a common protocol ( 36 ).…”

Real‐world evidence to support regulatory decision making: New or expanded medical product indications

“…28,29 This approach may be useful for providing a control arm for single arm trials, supplementing an existing control arm that is too small, or for providing an additional control arm in trials where the selected control arm is outdated due to a rapidly changing standard of care. 30 Control patients from RWD could be patients with indications for the medication under study who are receiving either an alternative treatment or no treatment if there is no currently indicated therapy. In either case, the use of RWD is intended to characterize the natural history, progression of disease, and patient outcomes in the absence of the medication under study to provide a control or "benchmark" for that observed on the experimental treatment.…”

Demonstrating Benefit-Risk Profiles of Novel Therapeutic Strategies in Kidney Transplantation: Opportunities and Challenges of Real-World Evidence

“…EC studies have different approaches in utilizing RWD for contextualization of trial data, and to supplement single arm trials. ECs can be used independently, for further contextualization while having a control arm in an RCT, or to supplement a control arm in an RCT ( 28 , 29 ). ECs sourced from RW settings reflect the SoC, and whilst finding these control cohorts can be challenging and resource intensive, they provide context to the benefits and risks observed in single arm studies, and can provide insight into RW patient experiences.…”

“…However, the practical considerations when using federated data models, such as ensuring linkage of disparate data sources, warrant caution ( 31 ). Use of RWD cohorts in innovative trial designs need to be aligned to prospective single arm trials with regards to population characteristics and definitions of data collected ( 28 ). Thus, to maximise the utility of harmonization by robust linkage and comparability, registries should more proactively develop common data modes to enable future research ( 39 ).…”