The power of natural history studies in rare disease R&D

Natural history studies are central to rare disease research. Rare diseases are defined as conditions that affect less than 1 in 200,000 people in the US or less than 1 in 2,000 people in Europe. With small patient populations that can be hard to reach, traditional clinical trial models are insufficient for rare disease drug development. 

Since natural history studies remove the burden of participation for patients and rely on observational data from medical records across geographical regions, they can aggregate a significant number of patients. In this blog, we explore how longitudinal natural history data can accelerate rare disease drug development and highlight successful use cases in recent years.

Natural history data points and longitudinal studies

Natural history data typically include medical diagnoses, treatments, clinical observations, and patient-reported measures (e.g., quality of life). Since most rare diseases are correlated with genetic changes, some natural history studies include genetic testing. This can help identify biomarkers that may be associated with prognosis or response to treatment, allowing for better patient stratification. 

Longitudinal studies track patients over time, therefore producing a rich set of information on disease evolution. In this manner, these studies can help map disease trajectories without medical intervention. This benchmark is highly valuable for clinical trial design, as it can provide an external control arm. That is, researchers can compare outcomes between patients who receive standard of care or no treatment at all ("control" groups) with those who receive new therapies that are being developed.

Impact of natural history data in drug development

Natural history studies can serve multiple functions in the drug development process for rare disease.

• Evaluate and refine how outcomes are measured, ensuring that they reflect meaningful changes in how patients feel and function under treatment. Assessments can include clinician-reported outcomes, observer-reported outcomes (e.g., caregiver reports), patient-reported outcomes, and performance outcomes (e.g., physical ability tests). 
• Identifying the patient population. Such studies often include genetic and phenotypic variations of diseases. This granularity helps researchers pinpoint who might benefit most from new drugs, thus optimizing clinical trial design parameters such as inclusion criteria, disease stage, trial duration, data collection schedules, and endpoints.
• Identifying and developing biomarkers that can help in diagnosis and prognosis. 
• Designing externally-controlled studies. By providing historical control data, natural history studies offer a baseline against which the effects of new treatments can be measured, helping to establish their efficacy and safety. This approach can also help sponsors address the ethical challenge of withholding treatment from participants.

In recognition of their importance, the FDA actively funds natural history studies that can fill knowledge gaps in rare disease drug development. 

Notable examples

In a recent natural history study on X-linked dystonia parkinsonism, 87 men were tracked over 18 months. Genetic testing was included and allowed researchers to identify changes based on genetic variants. Importantly, the investigators pinpointed a minimal battery of 21 measurements that were effective, non-invasive, and inexpensive that could be used in future trials, in addition to a relevant endpoint.

Another example is the Duchenne Natural History Study, the largest of its kind in Duchenne muscular dystrophy (DMD) thus far. It spanned over ten years and involved 440 participants across nine countries. Data from this study led to the discovery that longer term glucocorticoid treatment (for at least 1 year) reduced the risk of disease progression and even death. The investigators also identified a set of 9 clinical milestones that were highly predictive of disease trajectory. These findings can contribute to better trial design and enhanced patient stratification.

Future perspectives

One major advance is the use of AI to process rich longitudinal natural history datasets. Machine learning techniques can help reveal hidden patterns in disease progression, stratify patients by risk, and identify novel prognostic biomarkers, even when sample sizes are small or data are noisy. This kind of modeling can support the creation of synthetic or external control arms, reduce placebo use, and optimize trial designs.

Another key direction is standardizing prospective natural history studies to facilitate pooling of data across regions and countries. This will also improve uniformity and data integrity, enabling researchers to plan trials with more confidence. These efforts to build and sustain high-quality natural history studies will help accelerate rare disease drug development. 

If you’re interested in learning more about how to harness natural history for better rare disease trial design, download our whitepaper Foundations of clinical innovation: How insights from natural history studies drive drug development.