Clustering of Rare Medical Conditions Based on Clinical Features Which Determine Applicability of Investigative Designs and Methods to Their Study

Background:
The EU legislation determines that market access to new drugs requires the same level of evidence regardless of whether they are intended for rare or highly prevalent diseases, but the reality is that regulators often have to take decisions on orphan medicinal products (OMP) based on limited amounts of evidence. Clinical development for new orphan drugs with a cost-efficient design and analysis, an optimal fit to needs of patients, and reliable and efficient to decision making is thus of utmost importance to patients, regulators, and the pharmaceutical industry. Within an international FP7 collaboration (ASTERIX project), we have established a systematic and clinically based clustering in order to propose designs and analyses for trials conducted in small populations.
Material and methods:
A dictionary of 76 characteristics that are relevant to study design was created, and 27 diverse rare medical conditions were analyzed qualitatively for each characteristic. A database was created for this purpose, and a Multiple Correspondence Analysis (MCA) was performed to guide a first proposal of clusters, which has been further validated with expert clinicians and regulators.
Results:
A clustering of clinical conditions has been proposed based on MCA of 63 of the 76 potential characteristics. Relevant dimensions explaining data were identified, and 5 potential clusters of conditions were proposed based on common methodological approaches applicable to be conducted in drug development for each cluster. A sixth group of disease conditions was identified through clinical validation of groups. The time course and prognosis of the condition, as well as the type of end points, were the clinical features that most influenced the clustering.
Conclusions:
Six clusters of conditions to which potentially different clinical trial methodologies may be applied have been proposed. Further validation within the ASTERIX project of the usefulness of these clusters for methodological decision during drug development is ongoing.