AIMS: To evaluate the robustness of circulating tumor DNA minimal residual disease (ctDNA-MRD) test results to common factors that could impact performance. METHODS: DNA from three sample types is required: somatic, germline, and MRD monitoring. Different sample sources were evaluated for germline whole blood, peripheral blood mononuclear cells (PBMCs) and somatic (pre-treatment plasma, tumor) samples. Extraction methods, common interfering substances, and DNA input mass were evaluated for impact on MRD determination. Each comparison was performed in isolation. Overall test variability was also evaluated. RESULTS: Whole blood and PBMCs were acceptable germline sample types, with minimal differences in identification of somatic variants. Both pre-treatment plasma and tumor were acceptable somatic sample types, achieving the same 95% hit rate of three mutant molecules. Agreement between results from different extraction methods and in the presence of interfering substances was 100%. Positive percent agreement (PPA) between results from DNA input masses above and below the test upper and lower limits decreased, though PPA was 100% between 4 ng and the lower limit (5 ng). Variance in test performance fell within acceptable ranges (coefficient of variation 4.95%-10.33%). CONCLUSIONS: Major potential sources of variation and interference did not significantly impact ctDNA-MRD test performance.
Klimova et al. (Wed,) studied this question.