Natural history studies (NHS) can support regulatory decision-making at different stages of the drug product life cycle and are especially important in the context of rare diseases, which are associated with not only delayed or erroneous diagnoses but also a lack of approved treatments. Real-world evidence can fill knowledge gaps and support treatment decision-making, thereby benefiting affected patients. In this context, there are three important options for NHS design: retrospective, prospective, and cross-sectional. Each of these has been successfully used to support regulatory approval as external comparator arms (ECAs) for clinical studies, especially single-arm trials (SATs). While longitudinal data obtained from retrospective or prospective designs have been more commonly used and have been the focus of regulatory guidance documents, hybrid designs that combine retrospective and prospective data collection are particularly powerful for rare disease studies. This is due, in part, to the smaller number of patients impacted by each rare disease. In these settings, retrospective or prospective data collection alone may not be sufficient or fit-for-purpose for an external comparator. Rather, a strategic combination of all available data, regardless of timing, can deliver the right information of the desired quality and completeness to answer these important questions and support regulatory evidentiary needs. For instance, patients included in retrospective studies may differ from recently treated patients in terms of disease severity, disease variants, clinical management, or other important aspects of the disease that may impact patient outcomes. Further, retrospectively collected data may lack specific data elements required to achieve adequate comparison with the treated group in single-arm studies. In the context of prospective designs, the recruitment of sufficient new patients for prospective follow-up may not be feasible or may be prolonged due to the rarity of the disease. Further, the potential for premature truncation of patient follow-up may result in insufficient longitudinal data, or prospectively collected data alone may not provide insights into the disease course for specific groups of patients. In these situations, primary data collection in a prospective study may be supplemented with retrospectively collected data from chart reviews, registries, or electronic medical record databases, either for the same patients, in an ambispective design, or for a different set of patients. These hybrid designs allow for broader and more robust contextual information on the patient journey and the natural course of the disease to be obtained, which can improve the suitability of the data as an external comparator for SATs or studies that lack internal control in situations where a prospective design alone might not be sufficient. Because retrospective and prospective data, or any two data sources that are being combined, may differ in availability and quality, there are unique challenges alongside the strengths of these designs. In this paper, we discuss considerations for the design, analysis, and conduct of hybrid NHS intended as ECAs for single-arm studies in clinical development programs for rare diseases.
Read full abstract