From constraint to catalyst: innovation in rare disease clinical development

Rare diseases highlight some of the deepest structural challenges in clinical development. Patient populations are small, data is scarce and traditional randomised controlled trial models are difficult to implement. Patients with rare diseases, often facing few or no treatment options, are naturally reluctant to join traditional randomised trials where they might only receive a placebo or standard of care.

At the same time, regulatory thinking is evolving. Most recently, the FDA has announced it will allow sponsors, in certain cases, to file for drug approvals with just one pivotal trial rather than the traditional requirement for two. This move reflects growing flexibility in development models and changing expectations around what constitutes regulatory grade evidence, something rare disease programmes have been navigating for years.

In many ways, rare disease development acts as a laboratory for the future of clinical research. When patient numbers are limited and every data point matters, development teams are forced to find more efficient ways to generate evidence, reduce uncertainty and accelerate decision-making. These constraints, while challenging, have driven innovation in methodology and data use, offering valuable lessons to guide future clinical development strategies.

Small populations, big challenges

Rare indications frequently lack the depth and scale of data required to support traditional trial designs. Many conditions affect fewer than one in 100,000 people and symptoms, disease progression and treatment response can vary significantly across patients. In addition, rare disease trials typically involve children, since these conditions often present early in life, which further complicates the feasibility of conducting randomised controlled trials.

Many conditions affect fewer than one in 100,000 people and symptoms, disease progression and treatment response can vary significantly across patients.

This scarcity of patients creates a cascade of challenges for development teams. Recruitment is often slow, with patients dispersed across geographies and healthcare systems. Historical datasets are limited or fragmented, making it difficult to build a robust understanding of natural disease progression. Establishing clear endpoints, dosing strategies or comparator arms becomes more complex when baseline knowledge is incomplete.

In ultra-rare conditions, the scale of the challenge becomes even clearer. In some programmes, only a few hundred patients may exist globally, meaning that a clinical study would need to draw insights from a substantial proportion of the entire known population.

For many families, long diagnostic journeys and the absence of effective therapies remain a defining reality.

The result is a persistent tension between the need for rigorous regulatory grade evidence and the practical realities of studying diseases where patients are few, globally dispersed and often facing limited treatment options. For many families, long diagnostic journeys and the absence of effective therapies remain a defining reality. Against this backdrop, there is a clear need to rethink how evidence is generated and how trials are designed.

Reimagining trial design with clinical data science

Advances in clinical data science powered by AI and other big data technologies are beginning to change how constraints in rare disease trials are addressed. Techniques such as Digital Patient Profiles (DPPs) and digital twins allow researchers to build comprehensive representations of patient populations and the natural history of diseases, using curated real-world data and historical evidence.

Unlike anecdotal observations or generative outputs, DPPs are grounded in structured and contextualised clinical evidence and real-world patient data, both at individual patient level and at cohort level. By assembling contextualised data about patients, investigator sites and outcomes from large global datasets, DPPs form the foundation for building digital twins.

In rare disease trials, DPPs and digital twins are emerging as powerful tools to support development teams proven by growing results across four key areas:

1. Optimising trial design – Digital twins can simulate disease progression and treatment responses, allowing sponsors to refine inclusion/exclusion criteria, identify appropriate endpoints and select investigators with greater precision. Simply put, it allows the development teams to “meet the patients” even at the early stages of programme and trial planning. By evaluating trial protocols early, sponsors reduce the risk of costly amendments later.

2. Reducing external control arms – In rare diseases, placebo or comparator arms are often impractical or ethically challenging. Digital twins can be used to construct external control arms leveraging real-world datasets, minimising the need for traditional control arms while still supporting robust analysis.

3. Efficient use of limited patients – When only small cohorts exist globally, it is critical to maximise the insight gained from each participant. Digital modelling ensures that trials are designed to extract the most meaningful data while minimising unnecessary exposure to ineffective treatments.

4. Accelerate early-stage development – Advances in AI mean creating DPPs and digital twins is significantly faster and more accessible than in the past. These innovations allow development teams to build data foundations earlier in the process, informing study design and strategy before clinical trials begin. It can also allow for a more accurate interpretation of early results.

Importantly, advances in AI and clinical data science are helping teams extract insight from fragmented and unstructured data sources such as publications, clinical datasets and patient records. This enables a deeper understanding of rare disease populations across geographies where data may otherwise remain siloed, while also supporting precise trial design and execution. AI-first biopharma companies have shown that embedding AI into the core of operations can accelerate clinical trials by up to 20 percent.

However, technology must not be seen as a substitute for human expertise. AI can augment and accelerate data analysis and aggregation, but expert interpretation, validation and clinical and medical judgement remain essential. Ensuring that findings are meaningful, reliable and clinically relevant requires a combination of machine analytics and human expertise.

Regulation, evidence and the patient voice

As technology capabilities evolve, regulatory thinking is also adapting. There is increasing openness to alternative evidence approaches such as external control arms, digital modelling and other methods designed to address the limitations of small patient populations. The FDA has demonstrated this by reviewing the use of digital twins to generate simulated clinical records in trials as part of its evolving regulatory framework.

There is increasing openness to alternative evidence approaches such as external control arms, digital modelling and other methods designed to address the limitations of small patient populations.

Regulators, however, continue to emphasise the importance of transparency, robust data provenance and early dialogue between sponsors and agencies. Establishing alignment on what constitutes regulatory grade evidence is essential, particularly when novel approaches are being introduced.

While the clinical development industry will certainly reach a point where external controls are a regular part of the toolkit, the recent drug failure for Huntington’s disease shows that there is no one-size-fits-all solution. In diseases that are highly heterogeneous or where there is a strong expectation in patients of a curative effect, alternative forms of evidence may still be required.

The FDA has demonstrated this by reviewing the use of digital twins to generate simulated clinical records in trials as part of its evolving regulatory framework.

Alongside increasing openness from regulators, patient-centricity has become a central element of rare disease development. Clinicians may be experts in clinical science, but people living with rare conditions and their caregivers bring critical perspectives into how diseases affect daily life and what meaningful treatment outcomes look like.

Incorporating patient perspectives into trial design shapes more relevant endpoints and ensures that studies capture outcomes that matter not only to regulators, but also to patients and families. To achieve this requires going beyond relational engagement to structurally patient-centred engagement. Proactively building relationships with patient associations is an important first step, but now sponsors must focus on embedding patient insights into study design, endpoints and development strategy. Contextualised clinical data collected from all relevant patients is an essential part of that effort.

Lessons for the wider industry

Rare disease programmes have deep experience of some of the most complex challenges clinical development comes up against: limited patients, fragmented datasets and high scientific uncertainty. Yet these very limitations have driven some of the most innovative approaches in clinical development.

Advances in analytics, AI and data integration are helping researchers extract deeper insight from the patients and evidence that do exist. This approach supports more precise trial design and accelerates cycle times, helping reduce delays in bringing therapies to those who need them most.

At the same time, progress depends on maintaining rigorous data standards, strong regulatory engagement and genuine collaboration with patient communities. Innovation must be balanced with reliability, ensuring that new approaches are scientifically sound, regulatory grade and patient-centric.

Ultimately, rare disease research demonstrates that when evidence is scarce, innovation in data, methodology and collaboration becomes essential. The lessons learned in this space are not limited to rare conditions alone. As the wider pharmaceutical industry continues to explore more flexible, data-driven approaches to clinical development, many of the strategies pioneered in rare disease research may transform the future of how new therapies are brought to patients.