Technology Assessment on Genetic Testing or Molecular Pathology Testing of Cancers with Unknown Primary Site to Determine Origin

摘要: Objective This technology assessment reports the results of our review existing literature on commercially available genetic tests that are used to identify tissue origin (TOO) cancer in patients with unknown primary (CUP) site. CUP is a case metastatic tumor for which TOO remains unidentified after comprehensive clinical and pathologic evaluation. focused analytical validity their utility guiding diagnosis treatment improving health outcomes. Data Sources The scope was limited United States. We identified or molecular by searching GeneTests.org , NIH Genetic Testing Registry, GAPP Knowledge Base, following Food Drug Administration databases: Premarket Notifications (510(k)), Approvals, Clinical Laboratory Improvement Amendments. conducted searches PubMed, Embase, Cochrane Library. also searched Internet, once were identified, we grey search manufacturer’s Web sites. Review Methods included systematic reviews, randomized controlled trials, nonrandomized prospective retrospective cohort studies, case-control series published from 1990 present. excluded non-English studies; preliminary found very few studies other languages. relevant but did not contact authors additional data. conference presentations posters when they presented data elsewhere. Studies rated methodological quality. synthesized across each test using meta-analytic approach appropriate. Results reviewed cytogenetic analysis three genomic (CancerTypeID, miRview, PathworkDx) validity, utility. evidence these areas variable. Some analytic performance all tests, sufficient confirm only PathworkDX test. could compare because different reported assess statistical algorithms CancerTypeID miRview. unable algorithm PathworkDx TOO. Each has more publications report accuracy identifying known rates fairly consistent. summary 95% CI as follows: CancerTypeID—85 percent (83% 86%); miRview mets—85 87%), PathworkDx—88 (86% 89%). cases easily determined, actual most cases. contributed moderate. Low supported clincal usefullness making decisions. length survival amnog who received insufficient answer key questions effect Conclusions similar, ranging 85 88 percent. contribute do have decision Future Research Most current funded wholly partially manufacturers tests. urgent need evaluated research groups no evident conflict interest. Given difficulty assessing cases, future should focus benefits patient terms decisions resulting These will help value