The European Medicines Agency (EMA) has recently published its long-awaited final reflection paper on the use of single-arm trials (SATs) in drug development. As expected, and in typical EMA fashion, the agency’s opening salvo doesn’t mince words: “Randomised controlled trials (RCTs) are the standard for providing confirmatory evidence on the efficacy of an investigational treatment.”
Translation? Anything less than an RCT is going to be scrutinised to within an inch of its life. The agency is clear: single arm studies may have their place, but submitting one as pivotal evidence means you need to justify your decision comprehensively. Development teams should anticipate an uphill battle in an environment where single arm studies are inherently perceived as having fallen short of the methodological gold standard.
For drug developers grappling considering a single arm trial as a core part of their submission package the agency's reflection paper is a playbook of both caution and opportunity. Below, we unpack the key takeaways and what they mean for your asset.
RCTs Remain King—anything else is an Exception
The EMA’s stance couldn’t be clearer: RCTs are the gold standard, while single armed trials are the unconventional underdog. Submitting a single arm study as pivotal evidence will require a rock-solid explanation for why randomisation was impossible. Rare diseases, where patient recruitment is a challenge, might justify the choice, but this argument needs to be water-tight and backed by robust, verifiable evidence.
Endpoints: The Make-or-Break Factor
Endpoints in single arm studies bear the heavy burden of isolating treatment effects. The EMA emphasises that endpoints must be clinically meaningful and capable of showing clear efficacy. Binary endpoints (e.g., “cured vs not cured”) fare better than continuous or time-to-event measures, which can be confounded by disease variability or natural progression.
Counterfactuals Are Kingmakers
Without a control group, regulators will interrogate your assumptions about what would have happened without treatment. This “counterfactual” scenario relies on robust external data or natural history studies. Any uncertainty here can sink your efficacy claim. The importance of a high quality real world evidence study to support any treatment effect claim can't be overstated.
Pre-Specify or Perish
The EMA warns that single armed trials are particularly vulnerable to "post hoc rationalisations". From inclusion criteria to success thresholds, every decision must be pre-specified, justified and documented. Any unplanned protocol amendment or weakly justified assumption will trigger raised eyebrows and a host of questions.
Statistical Rigor Is Non-Negotiable
Confidence intervals are a baseline requirement, serving as a fundamental tool to quantify the uncertainty around the estimated treatment effects. However, regulators expect developers to go beyond merely providing these intervals; they must also define clear thresholds for success that are both scientifically and clinically justified. This involves setting specific criteria that the treatment must meet to be considered effective, and these criteria should be established before the trial begins to avoid any post hoc rationalisations. Furthermore, developers are encouraged to adopt conservative statistical approaches to account for the absence of randomisation, which inherently increases the risk of bias and variability in the results. These approaches might include using more stringent significance levels or employing advanced statistical techniques to adjust for potential confounding factors. By doing so, developers can bolster the credibility of their findings and provide a more robust argument for the efficacy of their investigational treatment.
Bias Mitigation should be a full chapter in any CSR:
The agency effectively sees single armed trials as breeding grounds for bias, including selection bias, attrition bias, and regression to the mean. The EMA’s reflection paper calls for proactive measures like blinding outcome assessors and transparent documentation of missing data. It is essential that drug developers take a comprehensive and systematic approach to not only designing their single arm study, but also identifying each potential source of bias and justifying how it's been controlled or that it's effect doesn't diminish the totality of the evidence.
Target Population Justification Is Critical
The EMA will scrutinise your inclusion criteria, recruitment methods, and subgroup analyses for any sign of misalignment from real-world patients and expectations.
It's essential that each inclusion criteria and recruitment method is justified and scientifically defensible. Significant caution should also be applied to site recruitment distributions as these will likely be heavily scrutinised.
Engage Early with Regulators
The EMA strongly encourages developers to seek scientific advice before embarking on single arm studies. Early dialogue can help fine-tune trial design and ensure regulatory expectations are met. It can't be overstated that early regulatory engagement always reduces risk when contemplating innovative or non traditional development strategies.
Be Ready to Defend External Validity
Single armed studies are often criticised regarding external validity, and the EMA is clear that results must be applicable to the broader patient population. Once again the role of principled and well conducted RWE studies can't be overstated here.
Avoid an unplanned “Phase II-to-Pivotal” strategy
Using exploratory single armed trials as pivotal evidence is risky. Unless you can demonstrate that your Phase II data meets the rigorous standards set of confirmatory trials, regulators are unlikely to be convinced.
The EMA dedicates significant attention to bias—and for good reason. Without randomisation, Single armed studies are inherently prone to distortions that can skew efficacy results. Selection bias, for example, can occur if trial participants are unrepresentative of the broader patient population. Attrition bias—caused by uneven dropouts—can further muddy the waters. Even the natural variability of disease progression can introduce regression to the mean, where patients improve merely because they started at their worst.
Mitigating these risks requires careful planning. The EMA recommends strategies like blinding outcome assessors, ensuring robust data collection, and conducting sensitivity analyses to stress-test your conclusions.
Endpoints are the linchpin of Single armed trials. The EMA is particularly cautious about time-to-event endpoints (e.g., progression-free survival), which are difficult to interpret without a control group. Continuous endpoints are also problematic unless they can definitively isolate treatment effects.
Binary endpoints, however, offer an alternative In diseases where outcomes like “cure” or “response” are unambiguous, Single armed trials may still deliver convincing efficacy data. The key is to predefine these endpoints rigorously and ensure they are clinically meaningful.
Since single armed studies lack internal controls, external data becomes the crutch upon which they lean. Historical data, natural history studies, and real-world evidence must fill the void left by randomisation. But beware: the EMA warns that discrepancies between external datasets and trial populations can erode the credibility of your conclusions.
While Single armed trials eliminate the need for a control arm, they don’t absolve developers of the need for robust statistical power. The EMA expects single armed studies to include enough participants to deliver reliable efficacy estimates. Undersized trials risk inflated variability and diminished confidence in results.
The EMA’s reflection paper is both a cautionary tale and a roadmap for navigating the single armed trials. For developers working on rare diseases or pioneering therapies, single armed trials remain a viable—albeit challenging—pathway to regulatory approval. But success will require meticulous planning, rigorous justification, and proactive engagement with regulators.
With the right strategy, a single arm can still deliver a knockout punch. Just remember: the EMA will scrutinise every move, so tread carefully and come prepared.
At Inovia Bio we've experts in the use of single arm trials, external data and real world evidence to support regulatory decision making: If you're considering these approaches fill in the form below to download our single armed study toolkit that guides drug developers step wise through he considerations they need to make to maximise their probability of success.