Criteria For Use Of Omics-Based Predictors In Nci-Sponsored Clinical Trials

58 Background: High-throughput omics technologies (e.g., genomics, epigenomics, proteomics, metabolomics) offer exciting opportunities for new biological insights into cancer. The IOM report on translational omics defined omics as the study of related sets of biological molecules in a comprehensive fashion. (IOM (Institute of Medicine) 2012. Evolution of Translational Omics: Lessons Learned and the Path Forward. Washington, DC: The National Academic Press.) The promise of omics technologies has proven problematic to translate into clinically useful tests. Difficulty obtaining biospecimens, unrecognized preanalytical influences, and suboptimal assay analytical performance can lead to unreliable results and conflicting reports. Poor reporting of study details and limited access to data and computer code can thwart efforts to replicate published results or to detect flaws in study design and analysis methods.NCI held an interactive workshop for a wide variety of stakeholders to explore better approaches to omics-based test development and validation. This workshop heavily informed the ideas presented here. Recommendations are stated concisely, then explained.A checklist of items to consider when evaluating the evidence for clinical use of an omics-based predictor, including in a trial where it will guide therapy, is presented. It covers specimen and assay requirements, the predictor model development process, clinical study design and conduct, and regulatory, ethical, and legal issues. The list applies to any trial involving investigational use of an omics test that will alter the clinical management of patients. The criteria also largely apply to situations in which the test will be evaluated retrospectively on specimens collected from patients who were prospectively enrolled on clinical studies.The proposed checklist should serve as a useful guide to investigators planning to submit proposals for NCI-funded studies involving use of an omics-based test. Ideally, this checklist will be consulted in the assay planning and development phases so that the necessary evidence will have been collected in a well-documented fashion by the time definitive evaluation of the test is desired.