By Kohkan Shamsi, MD, PhD, Principal, RadMD and Rick Patt, MD, Principal, RadMD
Medical imaging in clinical trials has been used as a measure of efficacy for more than 20 years. It was recognized by the FDA (FDA's 1994 Points to Consider for Developing Medical Imaging Drug and Biologic Products and Draft Guidance 2010) that variability in evaluating trial images necessitated a more standardized, controlled and independent review process. This independent review ("blinded read") is meant to improve the quality of imaging data evaluations from clinical trials. The performance of the reader(s) is a, and sometimes the, critical component in establishing drug or device efficacy. In this review we will discuss factors that can influence reader performance and how to actively manage readers who are blinded to clinical data to reduce bias.
In the clinical practice of radiology, a radiologist's image interpretation is designed for individual patient management, and uses subjective terms such as "improved" or "worsening" to describe a disease process. We refer to the pharmaceutical practice of radiology, whereby readers use standardized image evaluation criteria developed to evaluate drug efficacy. And as the images they review are blinded, these readers are not influenced in their review by medico-legal or insurance issues because these images are not impact patient management. Clinical practice rarely meets the requirements to enable selection, measurement, and tracking of lesions, and may use non-standard terminology. Moreover, this standard of care clinical practice varies geographically both within the US and globally. In addition, the pharmaceutical practice of radiology, using standardized imaging efficacy criteria used in trials (eg. RECIST, IWG, etc), is not taught in radiology residency training programs.