On the comparability of medical image measurements in cancer clinical trials

Biomarkers detectable by imaging methods are increasingly used as surrogate endpoints to establish the severity of a disease to evaluate the efficacy of a treatment method or chemotherapy drug in cancer clinical trials. In therapeutic cancer clinical trials, the tumors are measured from medical images such as CT or MR at a baseline time point and the treatment response assessment is followed-up on a frequent basis at subsequent time intervals. In this paper, we will try to evaluate to which degree the measurements from medical images are comparable for clinical trials by specifying sources of uncertainties of image measurements in order to support both clinical trial investigators by giving them confidence in their interpretation of the measurements results and implementers to follow the best-practice guidelines to be taken into consideration to minimize sources of such variability.