FDA Suggests Alternates to Placebo Control for Pivotal Clinical Trials
[Thursday, March 28, 2019]

It is hard to conduct placebo-controlled clinical trials, particularly in rare diseases. FDA released a new guidance document last week that provides ways to use historical controls to reduce and possibly eliminate placebo controls in all clinical trials. All diseases have a natural course under available standard of care which can be charted and used as external control in a clinical trial. Basically, in cases where good information about the progression of a given disease is known, FDA contends that clinical trials can be conducted just with the active arms using the historical information as comparator. The background or baseline disease information can provide valuable insights into the demographic, genetic, environmental, and other variables (e.g., treatment modalities, concomitant medications) that correlate with the disease’s development and outcomes. The most common place to find historical information about a disease progression can be found in disease registries. In case disease registries are not available or of limited quality, information can be collected from medical records, publications, insurance databases and other sources. Retrospective information on disease progression depends on the quality of the sources. Most medical records capture incomplete data, have selection biases, and have large variabilities from patient to patient owing to several factors. Prospective Natural History studies can be designed in such cases to address the historical control data. FDA suggests that natural history studies be designed early enough to capture detailed diseases progression information. The natural history studies can be initiated even before a specific investigational treatment is available for clinical trials or is at the preclinical stage. The prospective natural history studies are akin to patient surveys and do not require IND or IRB reviews since no intervention is involved. However, the patient privacy and consent to use their data is still relevant. The natural history studies can be conducted by patient support groups trying to establish natural history data to help developers of treatments for a given disease, by academic or public funded organizations, and even by insurance and medical groups. With the availability of electronic records and other real world data, historical control data is easier to collect and analyze. Natural history studies and retrospective historical studies can both benefit for robust statistical analysis. Overall, the emphasis is on using alternate data as control for studies where placebo control is hard. FDA suggests that such historical controls or external controls could not only play an important role in clinical trials in rare diseases but also be useful in clinical trials for non-rare diseases. This guidance is perhaps most relevant to owners of vast amounts of real world data on variety of health conditions looking for ways to use their data.