Considerations for Implementation of Cancer Molecular Diagnostics Into Clinical Care

Abstract

Physicians have provided personalized care with as much precision as possible for several centuries. However, increasingly sophisticated understanding of the human genome and of cancer biology has permitted identification of genetic and phenotypic distinctions that might permit development of new tumor biomarker tests for risk categorization, screening, differential diagnosis, prognosis, prediction, and monitoring. Both commercial and academic laboratories are offering tests for single analytes, panels of tests of single analytes, multiparameter assays coalesced into a signature, and total genomic, transcriptomic, or proteomic analyses. However, the absence of a consistent regulatory environment has led to marketing of assays without proven analytic validity or clinical utility. U.S. Food and Drug Administration (FDA) approval or clearance does not necessarily imply that use of the test will improve patient outcomes, and FDA discretion to permit laboratory-developed tests results in unknown benefit, or harm, of others. In this regard, a “bad tumor marker is as bad as a bad drug.” Caveat emptor is not a satisfactory approach to delivering high-quality care. Rather, adoption of tumor biomarker tests should be based on high levels of evidence generated in scientifically rigorous studies that demonstrate both analytical validity and clinical utility. Doing so will ensure that clinicians and patients are confident that a tumor biomarker test is likely to improve their outcomes.