Real-World Use of a 1021-Gene Liquid Biopsy Assay
Key Takeaways
- A 1021-gene liquid biopsy assay identified at least one on-label actionable biomarker in 16.18% of patients with metastatic cancer.
- Among 145 paired plasma and tissue cases, concordance for actionable on-label candidates was 90.34%.
- The assay captured SNVs and indels, CNVs, fusions, plasma TMB, and MSI, while parallel plasma and white blood cell analysis supported clonal hematopoiesis filtering and confirmation of pathogenic or likely pathogenic germline variants in 11.26% of patients.
Investigators performed 1,193 plasma analyses in 1,110 unique patients between October 2023 and May 2025. At least one on-label actionable biomarker was found in 16.18% of cases, with additional clinically relevant findings beyond that group. Broad plasma profiling added genomic information in this metastatic cohort and remained complementary to tissue testing.
The cohort included patients with metastatic cancer undergoing real-world ctDNA next-generation sequencing in routine oncology practice. Blood collection used 10 mL whole-blood samples in Cell-Free DNA BCT tubes, and 30 ng of cfDNA underwent NGS testing. When tumor tissue was available, library construction used 100 ng of DNA from the most recent specimen when sufficient material was available.
Non-small cell lung, breast, pancreatic, and colorectal cancers were the most frequent tumor types, and 77 patients underwent repeat testing, with an average of 12.9 months between the first and second test and repeated-testing follow-up spanning 2 to 29 months.
The panel detected somatic and germline alterations across the main variant classes, including SNVs and indels, CNVs, and fusions. Among oncogenic variants, 88.00% were SNVs or small indels, 10.67% were CNVs, and 1.32% were fusions. Frequently altered genes included TP53, KRAS, PIK3CA, APC, and LRP1B. Breast cancer had the highest proportion of on-label biomarkers, while colorectal cancer often showed RAS alterations linked to anti-EGFR resistance.
Plasma and matched FFPE tissue were analyzed in parallel for 145 cases, allowing direct comparison of actionable findings. Concordance for actionable on-label candidates reached 90.34% in that subset. FFPE tissue identified more off-label, trial-related, and variants of uncertain significance than plasma, although major on-label targets were seen with both methods.
Tumor mutational burden and microsatellite instability were also assessed, and 8.65% of the full cohort met plasma-based TMB-high and/or MSI-high criteria that may support consideration of immune checkpoint inhibitor therapy. Plasma thresholds were greater than 16 mutations per megabase for TMB-high and an assay-specific threshold of at least 18% for MSI-high.
Parallel sequencing of cfDNA and leukocyte genomic DNA was used to filter clonal hematopoietic mutations and support interpretation of plasma findings. This white blood cell comparator also enabled identification of findings consistent with pathogenic or likely pathogenic germline variants in 11.26% of patients. The same assay workflow incorporated plasma TMB and MSI assessment alongside somatic and germline analysis.
The authors noted uneven tumor-type representation, limited clinical records, and missing systematic data on treatment line, treatment status, tumor burden, disease site, surgery timing, and therapy exposure. Overall, the results reflect real-world implementation and concordance rather than prospective outcome-driven utility.