NEW YORK (GenomeWeb) – OncoCyte shared data this week from the latest validation and optimization of its blood-based lung cancer test, showing that the test meets the expected sensitivity and specificity thresholds for the firm to launch commercially this year as planned.
OncoCyte’s test is designed as a as a confirmatory tool — used after a lung nodule is discovered via radiologic imaging to help pick out patients who can avoid further interventions like a tissue biopsy.
Last year, company researchers and collaborators at the Wistar Institute who initially developed the assay reported 90 percent sensitivity and 62 percent specificity in distinguishing patients with benign nodules from those who turned out to have malignant cancers.
OncoCyte exceeded those metrics in the new data presented this week by University of Pennsylvania professor Anil Vachani at the American Thoracic Society 2017 International Conference. According to the presentation, the company’s optimized predictive algorithm demonstrated 95 percent sensitivity and 73 percent specificity in 299 samples collected prospectively from 29 US sites.
For the newer results, OncoCyte refined its gene-expression algorithm to 15 of the most predictive markers from an original panel of 18 genes. The new approach also integrates clinical information on the size