Clinical trials for rare diseases operate under a set of constraints that make nearly every stage harder than it is for common conditions. With roughly 7,000 recognized rare diseases and the vast majority still lacking approved treatments, the challenge is not a lack of scientific ideas but a shortage of patients to test them on. A phase 3 trial for the rarest conditions (those affecting fewer than one in a million people) enrolls a median of roughly 20 participants, compared with hundreds or thousands in trials for widespread diseases.1PubMed Central. Does the low prevalence affect the sample size of interventional clinical trials of rare diseases? An analysis of data from the aggregate analysis of clinicaltrials.gov That single fact reshapes how trials are designed, how evidence is judged, and how drugs reach the people who need them.
Why Small Numbers Change Everything
In a trial for high blood pressure or diabetes, researchers can recruit thousands of participants, randomize them into treatment and placebo groups, and detect even modest differences in outcomes with strong statistical confidence. Rare disease trials rarely have that luxury. An analysis of ClinicalTrials.gov data found that phase 2 trials for the rarest diseases averaged about 16 participants, and even diseases affecting up to 5 in 10,000 people only brought the average phase 3 trial size to around 75 to 78 patients.2PubMed Central. Does the low prevalence affect the sample size of interventional clinical trials of rare diseases? An analysis of data from the aggregate analysis of clinicaltrials.gov Those numbers are often too small to detect the minimum clinically important difference using conventional statistical methods.3PubMed. Design and analysis features used in small population and rare disease trials: A targeted review
The consequences cascade. When sample sizes are small, trials cannot easily split participants into treatment and control arms without leaving one group too tiny to analyze. Randomization becomes harder to justify ethically when every patient assigned to placebo represents a significant fraction of the global patient population. And a single dropout or data anomaly can skew results in ways that would be negligible in a larger trial. This is not just a statistical inconvenience; it fundamentally shapes the kinds of evidence regulators, doctors, and payers have to work with when deciding whether a rare disease treatment is safe and effective.
Adaptive and Basket Designs
One response to the small-numbers problem has been rethinking trial architecture altogether. Traditional fixed designs, where patients are randomly assigned to groups and the study runs until a predetermined endpoint, work well when participants are plentiful. For rare diseases, researchers have increasingly turned to adaptive designs that allow the trial itself to evolve based on incoming data. A Bayesian adaptive approach, for instance, can shift more patients toward the treatment arm that is performing better as the trial progresses, maximizing the number of participants who benefit from the superior treatment while still generating usable evidence. Simulation studies of one such design showed that a much higher percentage of patients ended up on the more effective arm compared to a fixed randomized design, and statistical power improved substantially over older adaptive methods.4Computational Statistics & Data Analysis. A Bayesian adaptive design for clinical trials in rare diseases
Basket trials take a different tack. Instead of studying one treatment in one disease, a basket trial tests a single therapy across multiple diseases that share a common biological feature, such as a particular genetic mutation or molecular pathway. This lets researchers pool patients from several rare conditions under one protocol, boosting enrollment without diluting the scientific question. A systematic review of basket trials in rare diseases found that this approach offers a promising way to address the constraints of small, dispersed patient populations.5PubMed Central. Basket trials in rare diseases: a systematic review of current practices, methodological challenges, and future directions The trade-off is complexity: analyzing results across different diseases requires careful statistical methods to avoid lumping together conditions that respond differently to the same drug.
When a Control Group Is Not Feasible
Some rare disease trials skip the traditional control arm entirely. When only a handful of patients exist worldwide, assigning half of them to placebo can feel both ethically uncomfortable and scientifically wasteful. External control arms offer a workaround by comparing treated patients in the current trial against historical or real-world data from untreated patients. This approach has already supported successful regulatory submissions and health technology assessments.6PubMed Central. External control arms for rare diseases: building a body of supporting evidence
Synthetic controls push this concept further. Instead of simply comparing against a static historical dataset, researchers use data-driven predictions of how individual patients would have fared under standard treatment. A proof-of-concept study in spinal cord injury demonstrated that synthetic controls reduced variability between groups and produced results that did not significantly differ from what randomized controls would have yielded, while allowing more patients to receive the investigative treatment.7PubMed Central. Exploring synthetic controls in rare diseases with a proof of concept in spinal cord injury The obvious risk is bias: if the historical data come from a different era of care, a different population, or use different measurement tools, the comparison falls apart. That is why natural history studies are so valuable.
The Underappreciated Role of Natural History Studies
Before you can test whether a treatment works, you need to know what happens without it. For common diseases, decades of clinical observation provide that baseline. For many rare diseases, the natural course of the condition is poorly documented. Natural history studies track patients over time, recording how the disease progresses, what symptoms appear, and at what rate. These studies do far more than fill medical textbooks. They help researchers choose the right endpoints for a trial, define the patient population that stands to benefit most, determine how long a study needs to run, and identify gaps in care that a new therapy could address.8PubMed Central. Novel clinical trial design and analytic methods to tackle challenges in therapeutic development in rare diseases
Patient registries built during natural history studies serve a dual purpose. They become recruitment pipelines, connecting researchers directly with patients and families already engaged with the rare disease community. And the data they collect can later serve as the external control for single-arm treatment trials. In diseases so rare that a registry might contain most known patients in a country, these databases become infrastructure that the entire research ecosystem depends on.
Measuring What Matters to Patients
Choosing the right way to measure whether a treatment works is a challenge in any trial, but in rare diseases, the usual yardsticks often do not exist. For a common condition like heart failure, there are well-validated outcome measures built over decades of research with large patient groups. For a disease affecting a few hundred people worldwide, researchers sometimes have to build those tools from scratch. Patient-reported outcome measures, which capture how a person feels or functions in daily life, are increasingly recognized as essential. But developing a reliable, disease-specific questionnaire is time-intensive and resource-heavy, and the problem is amplified in rare diseases because the small, scattered patient populations make it hard to recruit enough people even for the validation studies needed to confirm the tool works properly.9PubMed Central. The Use of Patient-Reported Outcome Measures in Rare Diseases and Implications for Health Technology Assessment
Creative solutions are emerging. Item banking and computer-adaptive testing allow researchers to draw from pre-built question libraries and tailor assessments to each patient, reducing the burden of developing an entirely new instrument while still capturing outcomes that matter to people living with the disease.10PubMed Central. Patient reported outcome measures in rare diseases: a narrative review In Gaucher disease, for example, a disease-specific questionnaire was developed with both a version for routine clinical monitoring and a shorter version for clinical trials, and validation in 46 patients showed excellent internal consistency.11PubMed Central. Development and validation of Gaucher disease type 1 (GD1)-specific patient-reported outcome measures for clinical monitoring and for clinical trials These efforts illustrate both what is possible and how painstaking the work remains.
Surrogate endpoints add another dimension. Rather than waiting years to see whether a treatment prevents disability or death, regulators sometimes accept a biomarker, such as a blood test or imaging marker, as a stand-in for the clinical outcome. The FDA’s accelerated approval pathway was designed with this flexibility in mind, allowing drugs to reach patients based on surrogate evidence while requiring follow-up studies to confirm the benefit.12PubMed Central. Recommendations for the development of rare disease drugs using the accelerated approval pathway and for qualifying biomarkers as primary endpoints For rare diseases, where waiting for definitive clinical outcomes could take decades with tiny patient numbers, surrogates can mean the difference between a drug reaching the market and a development program being abandoned.
Regulatory Incentives and Their Limits
The Orphan Drug Act of 1983 remains the backbone of rare disease drug development in the United States. By offering tax credits, reduced regulatory fees, and seven years of market exclusivity, the law was designed to make developing treatments for diseases affecting fewer than 200,000 Americans financially viable. It worked, in the sense that orphan drug approvals surged. But a longitudinal analysis found that the market exclusivity incentive has played a diminishing role in protecting rare disease drugs from competition over time, even as orphan drugs have grown as a share of all new drug approvals.13PubMed. Evaluating The Impact Of The Orphan Drug Act’s Seven-Year Market Exclusivity Period The incentive structure got drugs started, but for many rare diseases, the fundamental scientific and logistical challenges remain unsolved. Most rare diseases still lack any approved treatment.14PubMed Central. Development of orphan drugs for rare diseases
Accelerated approvals bring their own complications. Drugs that reach the market on the strength of surrogate endpoints carry more uncertainty about their real-world benefits and costs, which creates friction downstream when health technology assessment bodies decide whether insurers should pay for them. An analysis of Canada’s drug review process found that negative reimbursement recommendations for rare disease drugs were most often driven by a lack of demonstrated clinical effectiveness or insufficient evidence, not by cost alone.15PubMed Central. Health technology assessment of drugs for rare diseases: insights, trends, and reasons for negative recommendations from the CADTH common drug review In other words, even when a drug gets regulatory approval, the thinness of the evidence base can prevent patients from actually accessing it through their health system.
Bringing Trials to Patients Instead of the Reverse
Geography is a quiet killer of rare disease research. When a condition affects a few hundred people spread across a continent, asking them all to travel repeatedly to a specialist center for trial visits creates a participation barrier that eliminates many potential enrollees. Decentralized trial designs flip this model. By using telemedicine visits, home-based sample collection, and remote monitoring through wearable devices and electronic assessments, these approaches reduce the burden on participants and keep more of them engaged. Evidence from the COVID-19 pandemic showed that decentralized approaches were more resilient to disruptions in enrollment and retention than traditional site-based designs.16PubMed Central. What role can decentralized trial designs play to improve rare disease studies?
Digital health tools also bring measurement advantages. A wearable sensor tracking mobility at home captures how a person actually moves in their daily environment, which can look very different from walking down a hospital corridor during a scheduled assessment. And digital platforms can more easily accommodate multiple languages and accessibility needs, opening trials to a more global and diverse patient pool.17PubMed Central. What role can decentralized trial designs play to improve rare disease studies? In one striking example, an online recruitment strategy based on a community registry enrolled 182 participants with dementia, including those with frontotemporal dementia, within ten months. Traditional methods at the same trial had recruited zero.18Communications Medicine. Digital health technology use in clinical trials of rare diseases: a systematic review
How Patient Groups Have Changed the Landscape
Rare disease patient advocacy groups have evolved far beyond their traditional role of providing emotional support and raising awareness. A survey of these organizations found that nearly 80% now engage directly in research activities, with the majority contributing to clinical trials, natural history studies, or patient registries.19PubMed Central. Emerging roles and opportunities for rare disease patient advocacy groups Close to half have initiated their own research programs. These groups play instrumental roles in recruitment, biobank creation, and registry maintenance, acting as connective tissue between scattered patients and the researchers trying to study their conditions.20PubMed Central. The involvement of rare disease patient organisations in therapeutic innovation across rare paediatric neurological conditions: a narrative review
Their influence extends into trial design itself. Patient input during the development of outcome measures helps ensure that the endpoints a trial uses actually reflect what living with the disease is like, rather than measuring something convenient for researchers but irrelevant to everyday life. Regulators have increasingly recognized this, and the push for patient-centered trial design in rare diseases has been stronger than in almost any other therapeutic area, partly because the patient community is so organized and vocal.
Finding Patients Who Do Not Know They Are Patients
Many rare diseases go undiagnosed or misdiagnosed for years, meaning that potential trial participants exist but are invisible to researchers. Machine learning applied to electronic health records is starting to change this. One approach used a cascade learning algorithm trained on “silver standard” labels, meaning patients provisionally flagged based on indirect evidence rather than confirmed diagnoses. In a case study focused on lipodystrophy, the algorithm outperformed existing prediction methods and identified previously undiagnosed patients in a national-scale health records database.21PubMed Central. Learning to Identify Rare Disease Patients from Electronic Health Records Other studies have used EHR-based algorithms to identify eligible trial participants or to manage randomization digitally, streamlining the recruitment process that is often the single biggest bottleneck in rare disease research.22Communications Medicine. Digital health technology use in clinical trials of rare diseases: a systematic review
Newborn screening programs represent another front. By testing for conditions at birth before symptoms appear, screening can identify patients at the earliest possible stage and create a direct path into clinical trials of early interventions. Programs like Early Check are designed explicitly to accelerate understanding of diseases, shorten the timeline for adding conditions to standard screening panels, and feed directly into research on new treatments.23PubMed Central. Early Check: translational science at the intersection of public health and newborn screening
Equity and Access Barriers
The difficulties of rare disease trials do not fall evenly. Patients face barriers including geographic isolation from trial sites, cognitive or physical limitations that make participation harder, socioeconomic constraints, and a diagnostic odyssey that can delay entry into any study by years. Minority groups are particularly affected by these obstacles. Research involving rare disease patients and experts has identified three major impediments to healthcare access for this population: inconsistent government regulations across countries, inadequate financial and treatment support networks, and insufficient information about rare diseases among general practitioners and the public.24Journal of Rare Diseases. Drug development for rare diseases: a case for patient-centricity, equity, and access to clinical trials
Global regulatory fragmentation compounds the problem. A treatment approved in the United States may not be available in Europe for years, or vice versa, because each agency has its own evidentiary requirements. Harmonizing approaches to data extrapolation, especially for pediatric populations, has been identified as a priority to speed access to new therapies while also limiting the number of children exposed to investigational treatments during trials.25PubMed Central. Regulatory strategies for rare diseases under current global regulatory statutes: a discussion with stakeholders Broader alignment on surrogate endpoints, digital health tools, and decentralized trial designs could make rare disease drug development more efficient and inclusive, particularly in settings where traditional trial infrastructure is limited or absent.26PubMed. Addressing global regulatory challenges in rare disease drug development
After Approval, the Evidence Gap Persists
Getting a rare disease drug approved is not the end of the evidence problem; in many ways, it is just a shift in who bears the uncertainty. Drugs approved through accelerated pathways carry more unknowns about long-term outcomes, and post-marketing requirements differ between regulatory bodies. A cross-sectional analysis comparing FDA and EMA oversight of orphan drugs approved between 2008 and 2018 found that roughly half of orphan drugs at each agency came with post-marketing requirements. The agencies differed in emphasis: the FDA focused primarily on post-marketing safety, while the EMA was more likely to require studies addressing both efficacy and safety.27Clinical Pharmacology & Therapeutics. Assessing Post-Marketing Requirements for Orphan Drugs: A Cross-Sectional Analysis of FDA and EMA Oversight
Health technology assessment bodies face their own version of this gap. Orphan drug manufacturers often cannot provide the kind of robust comparative effectiveness data that HTA agencies normally expect, because the pivotal trials were small, single-arm, or relied on surrogate endpoints.28PubMed Central. HTA Evidence in Rare Diseases: Just Rare or Also Special? Real-world evidence collected after approval has been proposed as one way to fill this gap, including arrangements where coverage is granted conditionally while further data are gathered.29PubMed Central. Real-world evidence for coverage determination of treatments for rare diseases But collecting real-world data on a disease that affects a few hundred people is, predictably, slow and expensive.
Gene Therapies and the Long-Term Monitoring Problem
Gene therapies represent perhaps the most transformative category of rare disease treatments, but they also pose a unique challenge for clinical evidence. Unlike a daily pill that a patient can stop taking if problems emerge, gene therapy delivers a permanent change. Once the therapy is administered, there is no way to withdraw it. That creates what researchers have called an ethical imperative to invest in long-term monitoring well beyond regulatory approval.30The Journal of Clinical Investigation. Assuring long-term safety of highly effective gene-modulating therapeutics for rare diseases
Detecting late toxicity is harder than it sounds. In conditions like spinal muscular atrophy, where neurodegeneration may begin before birth and before treatment starts, it can be nearly impossible to tell on clinical grounds alone whether a new symptom reflects the underlying disease unfolding over time or a delayed adverse effect of the therapy. Patient-to-patient variability in how the treatment distributes within the nervous system adds another layer of complexity. Children treated through newborn screening programs receive gene therapy earlier in life than the patients in the original clinical trials, so the long-term safety profile in these youngest patients is genuinely unknown territory.31The Journal of Clinical Investigation. Assuring long-term safety of highly effective gene-modulating therapeutics for rare diseases
N-of-1 Trials in Ultra-Rare Conditions
When a disease is so rare that even a small trial cannot recruit enough participants, some researchers have turned to N-of-1 designs, where a single patient serves as their own control by cycling between treatment and no treatment over multiple periods. A systematic review of N-of-1 studies in rare genetic neurodevelopmental disorders found twelve qualifying studies with an average of five participants each. The interventions targeted mainly neuropsychiatric symptoms, and a key strength was the use of personalized, clinically relevant outcomes. The main weakness was limited use of validated and generalizable measures, which makes it hard to draw broader conclusions from any single study.32Neurology. Systematic Review of N-of-1 Studies in Rare Genetic Neurodevelopmental Disorders: The Power of 1 N-of-1 trials will never replace larger studies for establishing treatment effects across a population, but for an individual patient with an ultra-rare condition and no approved alternatives, they can offer a rigorous way to determine whether a particular therapy helps that specific person. In a field where every patient counts, that has real value.