Background: Colvera™ is a test that detects circulating tumor-derived DNA in patients with colorectal cancer by assaying for the presence of methylated BCAT1 and IKZF1 in blood. This study describes the analytical and clinical performance characteristics of the test.
Methods: Validation was performed in accordance with ISO15189 and National Pathology Accreditation Advisory Council requirements. Spiked samples including 264 plasma and 120 buffer samples were randomized, divided into 8 batches of 48 samples, and processed over 8 days using 2 equipment lines (each line consisting of a QIAsymphony SP/AS, QIACube HT, and LC480); 2 reagent batches; and 2 operators to determine limit of detection, selectivity/specificity, precision, reproducibility, ruggedness, and susceptibility to commonly known interfering substances. Clinical performance was validated by assaying 222 archived plasma samples from subjects (n = 26 with cancer) enrolled in a previous prospective trial.
Results: The limit of detection for Colvera was 12.6 pg/mL (95% CI, 8.6-23.9 pg/mL), which equates to 2 diploid genome copies per milliliter plasma. No statistically significant difference was determined between testing days (n = 8), instrumentation, operators, or reagent batches in precision studies for the methylation-specific assays. The assay performance was unaffected by 9 commonly known interference substances, variations in bisulfite conversion, or quantitative PCR settings (cycling temperatures, incubation times, and oligonucleotide concentrations). For this clinical cohort, sensitivity and specificity estimates for Colvera were 73.1% (19 of 26; 95% CI, 52.2-88.4) and 89.3% (175 of 196; 95% CI, 84.1-93.2), respectively.
Conclusion: Colvera is a robust test and suitable for detection of circulating tumor-derived DNA by measuring levels of methylated BCAT1 and IKZF1 in human blood plasma.
© 2017 American Association for Clinical Chemistry.