Background/Objectives: The NETest is a blood-based, machine learning-enhanced multigene transcript assay designed to detect and monitor neuroendocrine tumors (NETs). This study evaluated the accuracy of the recently validated NETest2.0® (2025) to (1) detect the presence of disease and (2) assess its utility as a clinically meaningful tool for monitoring NET status across diverse patient cohorts, including post-surgical surveillance, observation (“watch-and-wait”), and treatment settings. Methods: This registry study (NCT02270567) evaluated two objectives. For Objective 1, 1290 samples from 886 patients, of which 404 had paired follow-up samples, were analyzed for concordance between NETest2.0® and imaging-detectable disease. For Objective 2, paired blood samples (n = 404; median interval 7 months IQR 4–13.8) from NET patients across specialized centers were assessed. NETest2.0® scores were correlated with clinically adjudicated disease status using imaging as the comparator. Cohorts included post-surgical residual disease detection (n = 71), post-surgical recurrence monitoring (n = 44), observation (n = 72), and treatment monitoring (n = 217; somatostatin analogs, PRRT, and other therapies). Analyses were performed by cohort and in aggregate. Results: For Objective 1, NETest2.0® (cut-off ≥ 50) demonstrated an AUC of 0.96, sensitivity of 91.9%, specificity of 94.9%, PPV of 98.4%, NPV of 77.1%, and overall accuracy of 92.5%. Performance was consistent across tumor grades and sites. For Objective 2, 286 patients (70.8%) were stable, and 118 (29.2%) had progression or recurrence. NETest2.0® score changes correlated significantly with outcomes: scores decreased in stable patients (median −14.6%) and increased in progressive disease (median + 15.4%; p 0%) in score was associated with progression. Diagnostic performance for detecting progression reached a sensitivity of 78.0%, specificity of 98.3%, PPV of 91.1%, NPV of 90.2%, and accuracy of 83.9%. Conclusions: NETest2.0® accurately detects disease and provides a clinically actionable tool for monitoring NETs. Its high specificity and predictive performance support risk-adapted surveillance, potentially reducing unnecessary imaging while identifying early progression across diverse clinical settings.
Gulati et al. (Fri,) studied this question.