Abstract
Background/Aim: Circulating tumor DNA (ctDNA) testing has emerged as a minimally invasive tool for precision oncology, enabling dynamic monitoring of tumor burden and treatment response. However, commercial ctDNA NGS assays often omit clinically important oncogenes, limiting accurate assessment of copy-number variation (CNV). Amplifications of MYC and MYCN are key drivers of tumor progression and therapeutic resistance, and their detection is required under the Korean National Health Insurance coverage criteria. We evaluated whether a custom spike-in panel added to the Avenio ctDNA Expanded Kit improves CNV detection for MYC and MYCN to meet these clinical and regulatory requirements.
Materials and Methods: Spike-in targets were designed with KAPA Target Enrichment Custom Designs and integrated into the Avenio panel. Reference materials (Horizon Structural Multiplex cfDNA Standard, 5% (MYCN ≈9.5 copies); Seraseq ctDNA Complete, 1% (MYC≈3.07 copies)) were measured in triplicate; Seraseq was additionally diluted 1:2 and 1:10. Eight cancer-free plasma samples established the baseline. Libraries were sequenced on a NextSeq 550Dx (high-output). CNV analysis used CNVkit v0.9.9 with custom parameters (reference spread threshold increased 1.0→1.5; GC upper limit relaxed 0.7→0.8, lower limit retained at 0.3). Log2 fold-change versus healthy controls assessed CNV signals.
Results: Mean exon coverage was 698.5 for MYCN (range=325.4-1081.2) and 740.3 for MYC (range=438.8-1221.7). In the Horizon material, all MYCN exons showed ≥3.6-fold change (mean 4.2; inferred CNV ≈8.2), concordant with expected amplification. Seraseq showed a mean MYC fold change of 1.46 (inferred CNV ≈2.94); diluted samples yielded CNV estimates of 2.79 (1:2) and 2.67 (1:10), indicating limited sensitivity below ~3 copies. One MYC exon reproducibly underperformed despite adequate coverage.
Conclusion: Incorporation of a spike-in panel into the Avenio ctDNA assay enabled reliable detection of high-level MYC/MYCN amplifications and fulfilled practical requirements for local reimbursement. The estimated CNV limit of detection in this setting is ≈3 copies. Further replicate testing and validation with clinical specimens are warranted to refine sensitivity and interlaboratory robustness.
- Circulating tumor DNA
- liquid biopsy
- copy number variation
- MYC
- MYCN
- next-generation sequencing
- spike-in panel
Introduction
The advancement of molecular diagnostics and next-generation sequencing (NGS) has ushered in the era of precision oncology. Among various diagnostic approaches, circulating tumor DNA (ctDNA) detection – a fraction of cell-free DNA (cfDNA) released from cancer cells – has proven valuable for monitoring cancer treatment. For example, Ito et al. recently demonstrated that a multigene liquid biopsy approach could detect emerging resistance to osimertinib in patients with EGFR-mutated lung adenocarcinoma, while Nakamura et al. reported dynamic changes in cfDNA biomarkers before and after ablative radiotherapy in patients with oligometastatic colorectal cancer, supporting the utility of liquid biopsy in metastatic settings (1, 2). In addition, Moldogazieva et al. comprehensively reviewed the genomic landscape of liquid biopsy in hepatocellular carcinoma, highlighting the expanding applicability of ctDNA-based technologies across diverse tumor types (3). Currently, the only FDA-approved ctDNA NGS tests are Guardant360 CDx (Guardant Health) and FoundationOne Liquid CDx (Foundation Medicine) (4). However, these laboratory-based tests face challenges related to cost and accessibility.
The Avenio ctDNA Expanded Kit (Roche Diagnostics, Basel, Switzerland) is an NGS-based ctDNA panel that targets 77 genes, including those listed in the U.S. National Comprehensive Cancer Network (NCCN) Guidelines and emerging cancer biomarkers. Its scalability makes it an excellent choice for clinical laboratories within hospitals that perform ctDNA testing in-house. While it is known for its reliable detection of single nucleotide variants (SNVs) and small insertions/deletions (5), its ability to analyze copy number variations (CNVs) is limited to a restricted set of genes (6). Notably, it does not include the MYC and MYCN genes, which play a crucial role in tumor growth regulation. Amplification of these oncogenes is associated with poor prognosis and treatment resistance (7). In South Korea, ctDNA tests covered by the National Health Insurance are required to include the MYC and MYCN genes (8). To address this limitation, a spike-in panel was constructed using KAPA Target Enrichment Custom Designs. While SNV detection using spike-in NGS panels is well-documented (5), CNV detection has not been extensively studied. This study aimed to evaluate the CNV detection capability of the Avenio ctDNA Expanded Kit modified with spike-in genes.
Materials and Methods
Study design and reference materials. CNV detection was assessed by measuring the reference material in triplicate: the Horizon Structural Multiplex cfDNA Reference Standard (5%) (MYCN, 9.5 copies), and the Seraseq ctDNA Complete Reference Materials (1%) (MYC, 3.07 copies), each measured in triplicate. To determine the limit of detection (LoD) for MYC, Seraseq underwent 1:2 and 1:10 dilutions. Baseline CNV analysis was established using plasma samples from eight cancer-free individuals. All samples were sequenced on the NextSeq 550Dx instrument using a high-output flow cell with the Avenio ctDNA Expanded Kit spike-in panel, and mean sequencing depth was used to define the baseline (Figure 1). This study was approved by the Institutional Review Board (IRB) of Seoul National University Bundang Hospital (IRB: B-2206-762-304).
Workflow of copy number variation (CNV) detection using the spike-in Avenio ctDNA expanded kit.
CNV analysis pipeline. CNV analysis was conducted using the CNVkit (version 0.9.9) with custom modifications. The reference spread threshold was increased from 1.0 to 1.5 to accommodate MYCN, preventing its exclusion due to strict reference constraints. Additionally, GC content filtering was adjusted, relaxing the upper limit from 0.7 to 0.8 while maintaining the lower limit at 0.3, thereby reducing GC bias for MYC and MYCN. These adjustments optimized the balance between signal sensitivity and background noise for high-GC regions typical of these oncogenes.
Comparative and statistical analysis. Since reference samples were measured once, traditional statistical approaches were unsuitable. Instead, log2 fold-change analysis was applied, comparing MYC and MYCN amplification to healthy controls. Fold-change values were calculated per exon, and averaged CNV estimates were used for inter-sample comparison. Coverage uniformity and detection thresholds were evaluated by assessing log2-transformed coverage ratios and comparing them to known copy numbers in reference materials.
Results
Sequencing performance. The mean sequencing coverage was 698.5 (range=325.4-1081.2) for MYCN and 740.3 (range=438.8-1221.7) for MYC (Figure 2). These values indicated adequate depth for CNV calling within the spike-in panel. Baseline control samples demonstrated stable coverage without significant inter-sample variation, confirming analytical consistency.
Coverage distribution of MYCN and MYC genes. (A) Coverage profile of MYCN (transcript: NM_005378.6) across the reference genome. (B) Coverage profile of MYC (transcript: NM_002467.6) across the reference genome. The x-axis represents the genomic coordinates of each gene, while the y-axis represents the sequencing coverage. The black line indicates the median coverage, and the shaded region represents the variability in coverage across samples.
MYCN copy number analysis. In the Horizon reference sample, all MYCN exons exhibited a fold change of at least 3.6, with an average fold change of 4.2 (corresponding to CNV = 8.2) (Figure 3A). This value closely matched the expected 9.5 copies, validating the quantification accuracy of the modified pipeline. The uniform elevation across exons further confirmed reliable amplification detection.
CNV estimation across MYCN and MYC exons in reference samples. (A) CNV values for MYCN exons in the Horizon reference sample. The red line represents CNV estimates from the Horizon reference, while violin plots depict the distribution of CNV values across all samples. The gray line represents CNV values in healthy controls. (B) CNV values for MYC exons in the Seraseq reference sample at different dilution levels (0.1, 0.5, and 1). The colored lines represent CNV estimates for each dilution condition, while violin plots illustrate CNV variability across all samples. The gray line represents CNV values in healthy controls. CNV: Copy number variation.
MYC copy number analysis and limit of detection. In the Seraseq reference sample, all MYC exons, except one, exhibited a fold change of at least 1.3, with an average of 1.46 (corresponding to a CNV value of 2.94) (Figure 3B). Following dilution, CNV values did not decline proportionally, measuring 2.79 after a 1:2 dilution and 2.67 after a 1:10 dilution, indicating potential sensitivity limitations. One exon (chromosome 8: 127,738,493-127,738,770) exhibited lower CNV detection despite comparable coverage, suggesting a constraint in the reference material design.
Estimation of detection limit. These findings suggest that CNV changes below 3 copies do not induce a sufficiently detectable shift in log2-transformed coverage ratios under these conditions. Given the absence of significant coverage variation in diluted Seraseq samples, the LoD for CNV detection in this setting is estimated to be approximately 3 copies. This aligns with previous reports indicating a detection limit of 2.3 copies for MET amplification (9), though differences in experimental conditions, such as the use of spike-in genes, may account for some discrepancies.
Discussion
This study demonstrates that modifying the Avenio ctDNA Expanded Kit with a spike-in panel enables reliable CNV detection for MYC and MYCN, addressing a practical limitation of commercial ctDNA NGS panels. The observed amplification levels were consistent with reference values, and dilution did not significantly affect CNV estimates, implying that subtle changes below the detection threshold might go undetected.
Previous studies have reported that CNVs are associated with treatment resistance, with certain genes exhibiting significant alterations in chemoresistant patients (10, 11). Moreover, CNVs have been shown to evolve over time during treatment, reflecting tumor progression (12). Given the clinical significance of CNVs in cancer genetics, accurate CNV detection in ctDNA is essential. For instance, Dobilas et al. demonstrated that preoperative ctDNA levels were significantly associated with poor overall survival in patients with ovarian cancer, suggesting that ctDNA analysis may serve not only as a monitoring tool but also as a prognostic indicator to guide treatment decisions (13). As regulatory requirements vary across regions, targeted modifications to existing NGS assays can enhance their adaptability to regional regulatory requirements and clinical practice, ensuring their clinical utility in diverse healthcare settings.
This study has certain limitations. First, we were unable to clearly determine the LoD for CNV analysis of spike-in genes due to insufficient repeated testing. Additionally, we did not validate the spike-in genes using clinical specimens. Further optimization of analytical pipelines could also improve sensitivity and accuracy.
Importantly, verifying the detection capability of spike-in genes offers a practical approach to evaluating and optimizing commercial ctDNA assays under real-world conditions. Since commercial NGS panels are designed within predefined performance parameters, integrating spike-in panels allows for systematic assessment and refinement of CNV detection. This strategy enhances the reliability of ctDNA analysis and expands its clinical applicability.
Conclusion
In conclusion, this study demonstrated the feasibility of incorporating spike-in panels to enhance ctDNA NGS assays. By integrating MYC and MYCN into the existing panel, we successfully met South Korea’s National Health Insurance criteria while also generating clinically relevant results to support patient treatment and prognosis assessment. Future research should validate this approach with patient samples, explore its applicability across sequencing platforms, and optimize analytical workflows to ensure consistent and accurate CNV detection.
Acknowledgements
We would like to express our gratitude to all the subjects for providing specimens for this study.
Footnotes
Conflicts of Interest
The Authors report there are no competing interests to declare.
Authors’ Contributions
H. Kim contributed to the design, data analysis, data interpretation, drafted and figure creation; B.M. Kim contributed to the data analysis; J. Kim contributed to the data analysis, data interpretation, and figure creation; S.H. Seo contributed to the conception, design, data acquisition, data analysis, data interpretation, and critically revised the manuscript; S.M. Hwang contributed to the conception, and critically revised the manuscript; K.U. Park contributed to the conception, and critically revised the manuscript. All Authors gave final approval and agreed to be accountable for all aspects of the work.
Funding
This work was supported by the Roche Diagnostics of Korea under Grant No. 2022-03423
Artificial intelligence (AI) Disclosure
No artificial intelligence (AI) tools, including large language models or machine learning software, were used in the preparation, analysis, or presentation of this manuscript.
- Received September 23, 2025.
- Revision received October 27, 2025.
- Accepted November 3, 2025.
- Copyright © 2026 The Author(s). Published by the International Institute of Anticancer Research.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.