BACKGROUND: Randomized clinical trials (RCTs) are considered as the gold standard study design minimizing the risk of bias of treatment effect. In some circumstances, especially when preliminary evidence demonstrates potential dramatic efficacy in well-defined small populations with high unmet needs, single-arm trials (SATs) may be accepted by regulatory agencies. With fast development of innovative therapies and curative potential of some drugs, SATs are widely discussed by decision-makers. What constitutes dramatic efficacy and high unmet medical need is debated with limited guidance for drug development and optimal management of remaining uncertainties. Draft guidance on natural history studies for drug development in rare diseases was recently launched by the US. Such studies may play a key role in clinical development programs of future pipeline based on SATs, as an external control approach.

DISCUSSION: HTA bodies have limited guidance regarding SATs. High uncertainty inherent to SATs are challenging drug reimbursement and pricing decisions and may compromise patient access. Use of SATs and external control approach should be enough justified to be accepted by HTA bodies: 1) Strong assessment and rationale for not doing RCT; 2) Preliminary data suggesting dramatic magnitude of treatment effect size versus historical cohort; 3) Treatment effect size objectively measured in a robust way minimizing bias; 4) External cohort relatively homogeneous and very similar to the treated group, or impact of study heterogeneity in the patient population and on outcomes to be studied; 5) Well-known confounding factors affecting outcomes and statistically sound matching method used to control for confounding; 6) Generalizability and transferability of clinical data toward historical cohort to be proactively assessed.

CONCLUSIONS: There is a need for continuous multi-stakeholder dialogues to define evidence requirements to satisfy regulatory bodies and HTA bodies/payers through operationalized guidelines for SATs and use of natural history data as an external control approach.