Sensitivity and Specificity of the NETest: A Validation Study

Background: Secretory tumor markers traditionally measured in patients with neuroendocrine tumors (NET) are lacking in sensitivity and specificity, and consequently of limited clinical utility. The NETest, a novel blood multigene RNA transcript assay, has been found to be highly sensitive and specific. We sought to validate the sensitivity of the NETest in a population of metastatic well-differentiated NETs of gastroenteropancreatic and lung origin and evaluate the specificity in a mixed population of metastatic non-NET gastrointestinal (GI) malignancies and healthy individuals. Design and methods: 49 patients with metastatic NETs, 21 patients with other metastatic gastrointestinal cancers, and 26 healthy individuals were enrolled. Samples were sent in a blinded fashion to a central laboratory and a NETest value of 0-13% was considered normal. Results: Using the upper limit of normal (ULN) of 13%, the sensitivity of the NETest was 98% (95% CI, 89% - 100%). The overall specificity was 66% (95% CI, 51% - 79%), with 16 false positive results. Specificity was 81% (95% CI, 62% - 92%) among 26 healthy individuals and 48% (95% CI, 26% - 70%) among patients with other GI malignancies. Using an updated normal range of 0-20%, sensitivity was unchanged, but specificity improved to 100% among healthy participants, and 67% among patients with other cancers. Conclusions: The sensitivity of the NETest is exceptionally high (>95%) in a population of metastatic, well-differentiated NETs. Specificity within a healthy population of patients is exceptionally high when using a normal range of 0-20% but relatively low when evaluating patients with other GI malignancies.