Circulating tumor DNA (ctDNA) testing is now a central component of precision oncology, enabling non-invasive tumor genotyping, detection of resistance mutations, and monitoring of treatment response or minimal residual disease ^1,2. While the clinical utility of ctDNA is well established in select indications—such as EGFR mutation testing in non-small cell lung cancer and ESR1 testing in breast cancer —broader application across tumor types has been driven not only by established indications but also by the advent of large, multi-gene panels³. These comprehensive assays, often marketed for pan-cancer use, are designed to capture a wide spectrum of genomic alterations and biomarkers, but their widespread adoption outpaced rigorous evaluation in real-world settings.

To support clinical implementation, laboratories are required under accreditation standards (e.g., CLIA, CAP, ISO 15189) to demonstrate not only the analytical validity of their assays, but also their clinical validity and utility ^4,5. In publicly funded institutions, demonstration of clinical utility is also essential for cost-justification. However, establishing clinical utility is inherently challenging in oncology, where the impact of genomic results on management decisions depends on a wide range of contextual clinical factors. As such, meaningful assessment requires robust clinical annotation, enabling linkage between assay results, therapeutic decisions, and outcomes. Furthermore, best practice guidelines for molecular test reporting—such as those issued by the Association for Molecular Pathology (AMP), American Society of Clinical Oncology (ASCO), and College of American Pathologists (CAP)—emphasize the integration of clinical context in variant interpretation and laboratory reporting ^6,7.

In this study, we evaluated a commercially available, high-cost, comprehensive ctDNA assay for both analytical performance and clinical utility across a large and diverse cohort of patients with solid tumors. The assay demonstrated strong analytical validity, with consistent detection of genomic alterations across multiple cancer types. However, efforts to assess clinical impact were undermined by inconsistent and often poor-quality clinical documentation, limiting the ability to determine whether and how results influenced clinical decision-making. These findings highlight the need for improved clinical data capture and integration as part of laboratory and hospital workflows to support regulatory requirements, optimize test utility, and ensure adherence to best practice standards in molecular diagnostics.

For analytical validity studies, we used commercial reference standards specifically designed for ctDNA assays with variant allele frequencies between 0.125% and 5% (0710 − 0141, 0710 − 0143, 0710 − 0144, SeraCare, covering a range of hotspot and non-hotspot mutations in 27 genes, as well as TPR::ALK and NCOA4::RET fusions; SID-000144, SensID, covering ESR1 hotspot mutations; HD786, Horizon, covering hotspots in GNA11, AKT1, PIK3CA, EGFR, as well as ROS::SLC34A2 and CCDC6 fusions). Reference standards were tested across three (0.5% and 0% standard) or four (0.125% standard) independent runs.

CtDNA testing of patient samples was performed on physician request in the course of clinical care. Blood samples were submitted to our institution using Roche Cell-Free DNA collection tubes (Roche, 07785666001). Plasma was prepared by centrifugation (1x 2000g for 20 minutes, 1x 3200g for 30 minutes). CfDNA was isolated from 10ml of plasma using EZ1&2 ccfDNA Kit (Qiagen, 9027011) according to manufacturer’s instructions, and frozen at -80°C until library preparation, typically within one week. Library preparation was performed using the AmoyDx Comprehensive Panel (AmoyDx), designed to detect small variants in 128 genes and fusions in 12 genes, according to the manufacturer’s instructions. In brief, the protocol comprises end-repair, adapter ligation, amplification (14 cycles), and several purification and quality control steps prior to target enrichment via hybrid capture, followed by additional amplification, purification and quality control (see below). Libraries were sequenced on a NextSeq550 or NextSeq500 using a mid-output flow cell and 2x150bp read configuration at a final library concentration of 1.3 pM.

Data processing from basecall (.bcl) files, including demultiplexing, was performed using the proprietary AmoyDx ANDAS Server (pipeline ADXPAN116 versions 0.2.0 und 0.2.1). Assessment of run performance involved control of sequencing quality (Phred Q ≥ 30, > 0.75), total clean data output (excellent: 6–11 gigabases, acceptable: 12–22), overall coverage (> 0.95), coverage at 1550x unique (unique molecular identifier – UMI – based) depth (excellent: >0.95, acceptable: >0.85). Precision was calculated as the mean deviation of individual allele frequencies from the mean allele frequency, divided by mean allele frequency. Variant annotation, interpretation and reporting was performed in accordance with Variant Interpretation for Cancer Consortium Standard Operating Procedure⁸ as part of a routine diagnostic workflow.

Approval of the ethical board of the Medical University of Vienna for the retrospective study was granted under 1989/2024 (informed consent was waived for retrospective use of data). Retrospective evaluation was performed for 536 tests based on 497 samples, processed from 12/2022 (after test implementation) to 06/2025. Analysis of diagnostic yield was restricted to 485 tests/samples with signed-out test results (i.e. excluding failed tests), representing 438 unique patients. Retrospective review of test requisition and electronic medical records was performed for the first 126 consecutive patients (without further exclusion criteria, from 12/2022 to 12/2023) by manual search of all entries from 1 months prior to sample submission until 2 months after reporting test results. Following data were collected: presence / legibility of diagnosis and explicit question on test requisition form, reasoning for preferring liquid-based over tissue-based testing, documented acknowledgment of results in any form, in particular in the context of molecular/multidisciplinary tumor board (MTB) documentation. A change in clinical management was attributed to ctDNA testing if it was documented in the medical record, was plausible based on the reported biomarker(s), and not previously identified by tissue-based testing. In contrast, in cases where tissue-based test results reporting the relevant biomarker(s) were issued prior to the ctDNA results (i.e., in case of parallel testing), the change in management was defined to have occurred due to tissue testing and not due to ctDNA testing.

For retrospective analysis, diseases were coded as entities according to MSKCC OncoTree⁹ as implemented in Cancer Genome Interpreter¹⁰. Automated assignment of clinical actionability was performed for all clinically reported oncogenic and likely oncogenic variants using Cancer Genome Interpreter (accessed via API on 11/07/2025), which provides clinical evidence tiers according to the VICC harmonized meta-knowledgebase¹¹ (based on ASCO/CAP/AMP recommendations⁶).

Graphs were generated using R Statistics¹² (4.4.3) and RStudio in combination with ggplot2¹³ and complexheatmaps¹⁴ packages. Language editing was performed using a large language model (OpenAI GPT-4o).

Assay Performance Characterization

To validate the performance of the assay, including the proprietary pipeline for variant calling and filtering, we conducted sensitivity and specificity studies using multiple reference standards across independent preparation and sequencing runs (Fig. 1A).

We first analyzed SeraSeq® cfDNA reference standards provided at variant allele frequencies (VAFs) of 0.5%, 0.125%, and 0% (negative control). At 0.5% VAF, the assay demonstrated analytical sensitivity of 98.3% for hotspot mutations and 100% for non-hotspot mutations. At 0.125% VAF, sensitivity decreased to 73.8% and 70%, respectively. Variants were detected with mean allele frequencies of 0.53% and 0.15%, indicating high trueness, with a bias of ± 0.03% VAF and absolute precision of 26.1% and 40.0% (corresponding to absolute VAF of 0.06 and 0.14%, respectively). Using negative controls (0% VAF), we established the limit of blank (LoB) at a VAF of 0.06%. However, some variant calls were observed at frequencies up to 0.15% (Fig. 1A).

Unexpectedly, the default AmoyDx variant filtering pipeline discarded many low-frequency variants, classifying them as invalid due to preset thresholds (Fig. 1A). This initially reduced analytical sensitivity (80% at 0.5% VAF), highlighting a critical limitation in the default pipeline that required custom filtering to unlock the assay’s full potential.

Further validation using additional reference standards designed to test ESR1 mutations (1% VAF) and a separate 5% VAF standard demonstrated 100% sensitivity and specificity (data not shown). Notably, the AmoyDx kit's positive control includes variants and fusions at VAFs ranging from 5–80%, which, although confirming high assay sensitivity (100%) and precision (9.5–1.8%, corresponding to absolute VAF of 0.51% and 1.41%) in this range (Fig. 1B), is inadequate for validating performance at lower, clinically relevant frequencies.

We next evaluated fusion detection (Fig. 1C). Of the two fusions present in the 0.5% VAF cfDNA reference, one was consistently detected across all three runs, while the other was detected in only one, yielding 66% sensitivity. No fusions were detected at 0.125% VAF. In contrast, all fusions in the 5% reference standard and kit positive control were detected at expected frequencies (3.3–20%), indicating high sensitivity at this range. We did not consider optimizing filtering thresholds for fusion detection because we observed recurrent detections of unexpected fusions involving FGFR3, LTBP1, NTRK1, and NTRK3 across all standards, suggesting potential false positives. NCBI BLAST analysis revealed that expected fusions from commercial controls aligned to both gene partners on standard RefSeq chromosomes. In contrast, potential false positive detections, but also fusions specified in the AmoyDx kit positive control, did not align, implicating the AmoyDx pipeline in the generation of spurious fusion calls and mandating caution when interpreting findings.

Technical Performance of the Assay in Clinical Samples

A total of 497 cfDNA samples were submitted for analysis. Due to failed minimum sequencing depth— defined as unique molecular depth exceeding 1550× in less than 85% of target regions—39 samples were re-tested from available cfDNA isolates, resulting in a total of 536 tests.

The median cfDNA concentration was 1.35 ng/µL, with a broad range of 0.12 to 56.4 ng/uL (Fig. 2A). Sequencing quality and overall target region coverage thresholds were met in virtually all tests (100% and 99.6%, respectively). However, excellent criteria for sequencing depth—defined as unique molecular depth exceeding 1550× in more than 95% of target bases—were met in only 367 of 536 tests (68.5%).

As expected, there was a statistically significant correlation between input cfDNA concentration and mean unique molecular (UMI-corrected) sequencing depth (Spearman’s ρ = 0.35, p < 0.0001; Fig. 2B). The likelihood of achieving the desired depth varied markedly by cfDNA concentration. Among samples with concentrations above 15 ng/µL, 75% reached the target depth. Similarly, the target was met in 78.3% of samples above 1.5 ng/µL, 77.5% above 0.75 ng/µL, and only 9.8% of those below 0.5 ng/µL. Notably, no sample with a cfDNA concentration below 0.25 ng/µL achieved the required sequencing depth (Fig. 2C). These findings suggest that while input concentration above 0.5 ng/µL is not a reliable predictor of adequate sequencing coverage, concentrations below this threshold are consistently associated with suboptimal performance and are unlikely to yield usable results.

Of the 497 samples tested, 485 final reports were issued (97%). Among these, 56.5% met excellent test performance criteria, 21.6% met acceptable criteria (as outlined in methods), and the remaining 21.6% did not meet the required overall criteria, but were reported with comments regarding poor test performance – e.g., in light of confident and pertinent individual findings.

Biomarker Yield and Mutation Landscape

A total of 485 reports were finalized, corresponding to 438 unique patients. Forty-three patients contributed multiple samples at different timepoints (see below). The patient cohort ranged in age from 25 to 94 years (median 66), with female patients accounting for 40% of the population. The most common tumor types were prostatic adenocarcinoma, breast cancer, non-small cell lung cancer (NSCLC), and colorectal adenocarcinoma.

Oncogenic mutations were identified in 346 of 485 samples (71.3%), with a mean of 1.6 mutations per sample (median 1; maximum 20), and were distributed across 60 of the 128 targeted genes, most frequently involving TP53, PIK3CA, KRAS, ESR1, and AR (Fig. 3A–B). Variants of unknown significance (VUS) were detected in 316 samples (74.2%), with a mean of 1.44 variants per sample (median 1; maximum 21).

To assess the potential clinical relevance of test results, we performed disease-specific biomarker annotation using the Cancer Genome Interpreter (Fig. 4A). Biomarkers with recognized clinical evidence (any tier) were identified in 319 samples (65.8%), while high-evidence, Tier A biomarkers were found in 131 samples (27.0%). In an additional 61 cases (12.6%), Tier A biomarker status was inferred based on pertinent negative findings (e.g., wildtype status for KRAS, NRAS, or EGFR). We verified that the inclusion of patients with multiple samples did not significantly skew overall biomarker yield. When restricting analysis to the first sample from each patient, the findings remained consistent: 69.9% of samples contained oncogenic mutations and 27.9% harbored Tier A biomarkers.

Regarding only positive findings, biomarker assertions were most frequently associated with detected mutations in PIK3CA, KRAS, ESR1, NF1, and PTEN (Tier A biomarkers: KRAS, ESR1, PIK3CA, BRCA2 and EGFR) (Fig. 4B–C). Across the panel, clinically actionable biomarkers—including both positive findings and pertinent negatives—were observed in only 43 of the 128 targeted genes. Only 16 genes were associated with Tier A biomarkers, suggesting that the breadth of the panel exceeds practical clinical needs in this setting.

When stratified by tumor type, breast cancer, lung cancer, and colorectal cancer accounted for the largest absolute number of cases with Tier A biomarker assertions, including both positive and negative findings (Fig. 4D–E). This distribution reflects both the relative frequency of these cancers in the cohort and their established molecular targets. In contrast, although prostate cancer was the most common single tumor type, only 15% of cases demonstrated Tier A biomarkers. By comparison, ≥ 50% of cases with NSCLC, colorectal adenocarcinoma, thymic carcinoma, and breast cancer showed high-evidence biomarkers.

Among the 43 patients with multiple tests, 40 submitted two samples, two submitted three, and one submitted four. The time interval between samples ranged from 4 to 783 days (median 290; mean 335) (Fig. 4F). Differences in biomarker findings were noted in 24 patients (56%, Fig. 4G). In 19 of these cases, a greater number of biomarkers and/or a higher tier of biomarker level was identified in a subsequent test (~ 80%).

Prostatic adenocarcinoma

Prostatic adenocarcinoma represented the most frequently tested cancer type in our cohort (Fig. 4D), accounting for 170 of the 457 samples (36%), which prompted us to characterize this group of patients separately. At our center, these patients predominantly represent cases of metastatic castration-resistant prostate cancer (mCRPC), typically following multiple lines of systemic therapy and often with limited tissue availability due to bone metastases.

Genomic alterations were most frequently identified in TP53, AR, PIK3CA, CDK12 and PTEN (Fig. 5A). Tier A biomarkers were identified in 15% of prostate cancer cases, the majority of which involved genes associated with homologous recombination repair (HRR), including CDK12, BRCA2, and BRCA1 (Fig. 5B). Notably, all observed CDK12 alterations were frameshift mutations, with more than half located in exon 1 (transcript: NM_016507.4), suggesting a potential regional mutational hotspot that may warrant further investigation.

In one patient with mCRPC, a high-frequency putative ROS1::STX8 fusion (involving exon 34 of ROS and exon 7 of STX, variant allele frequency ~ 30%) was detected in two independent plasma samples collected at different timepoints. Due to the absence of available tissue material, orthogonal validation by RNA sequencing was not possible. The patient did not have any other known or suspected malignancies. Unfortunately, the patient was lost to follow-up.

Retrospective Assessment of Clinical Utility

To evaluate the clinical utility of ctDNA testing in our institutional setting, we conducted a retrospective review of electronic health records for the first 126 consecutively tested patients. This analysis focused on the quality of information provided on test requisition forms, availability of relevant clinical context, acknowledgement of molecular results, and downstream impact on clinical decision-making.

A cancer diagnosis was clearly stated on 84% of test requisitions; in the remaining 16%, it was either absent or difficult to interpret (Fig. 6A). Only 24% of cases included a specific clinical question or suspected resistance mechanism to guide molecular analysis (Fig. 6B). Upon reviewing records from two months prior to and following sample submission, the rationale for ctDNA testing could not be confidently determined in 60% of cases. When present, the most frequent documented indications were previous insufficient tissue biopsy or anatomical challenges precluding re-biopsy (Fig. 6C). In three percent of cases, a formal tumor board recommendation was documented, albeit without an explicit rationale.

Despite the assay’s potential to guide management decisions, structured acknowledgment of test results was identified in only 40% of cases (Fig. 6D), and discussion during a multidisciplinary tumor board in 15% (Fig. 6E). Institutional variability was notable, with traceable indications ranging from 30–80% across departments (data not shown).

Clinical outcome data were also inconsistently recorded. ECOG performance status within one week of sampling was documented in only 40 patients (range 0–2, median 0). A change in patient management plausibly attributable to ctDNA results was identified in 11.7% of patients, with 5.5% receiving therapy targeting an identified actionable alteration and 6.2% undergoing guideline-concordant initiation of first-line treatment based on the ctDNA molecular profile (Fig. 6F).

Circulating tumor DNA analysis has emerged as a powerful diagnostic modality in oncology, offering minimally invasive, repeatable access to tumor genomic information. Its applications span from mutation detection and therapy selection to real-time monitoring of resistance evolution and minimal residual disease. In the context of solid tumors—particularly those with inaccessible or insufficient tissue—ctDNA profiling enables genomic insights that might otherwise be unavailable. However, the integration of this technology into clinical decision-making remains contingent not only on assay sensitivity and specificity, but on its thoughtful use within a clinically informed framework.

On support of current practice, this study reinforces the analytical validity of circulating tumor DNA (ctDNA) analysis and its capacity to identify clinically actionable biomarkers. The AmoyDx Comprehensive assay used in this study demonstrated high analytical reliability, with detection of known driver mutations and other variants with established therapeutic, prognostic, or investigational relevance in a significant subset of patients. These findings are aligned with a growing body of evidence supporting the sensitivity and specificity of ctDNA assays, especially in advanced or metastatic disease.

Previous prospective and retrospective studies have come to heterogeneous results when assessing clinical utility of ctDNA testing. The prospective PRISM study using FoundationOne Liquid CDx (324-gene panel, > 1000 patients) showed 94% assay success, 64% actionable alterations, and 21% receiving matched therapy³. In regards to retrospective data, a real-world analysis of over 3,000 patients with advanced NSCLC tested with Guardant360 showed actionable alterations in 41.9% of patients, and significantly improved survival with guided therapy (OS: 36.1 vs. 16.6 months; p < 0.001).¹⁵ Previous studies with smaller cohort sizes showed actionable alterations in only 20% (> 900 patients)¹⁶ or even merely 8% (199 patients)¹⁷. Much of these differences may be attributable to the inherent bias of “real-world evidence” studies ¹⁸, breadth and sensitivity of specific ctDNA assays, diversity in patient cohorts and even definition of target actionability.

In our study, much of the panel content remained underutilized, with oncogenic mutations detected in only 60 of 128 genes covered by the panel, and actionable alterations (as defined via CAP/ASCO/AMP Tier A-D) detected in only 16 (tier A) or 43 (tiers A-D) genes, respectively. Furthermore, it is plausible that many of these matching biomarkers were clinically irrelevant at time of testing due to advanced line of therapy. This suggests that optimization of panel content based on specific use-cases is necessary to achieve an ideal cost-effectiveness relationship. In our hands, the greatest cause of impaired or invalid test results was inadequate unique molecular coverage. This may be due to insufficient recovery of input molecules during library preparation, as demonstrated by frequently poor test performance despite adequate isolated cfDNA concentrations. Indeed, efficient recovery of rare DNA fragments from the initial sample has been argued to be one of the most critical steps in ctDNA analysis sensitivity¹⁹.

In the context of metastatic castration-resistant prostate cancer (mCRPC), where tissue biopsies are often infeasible or low-yield, and tumor evolution occurs rapidly under therapeutic pressure, ctDNA provides a non-invasive and dynamic window into tumor genomics²⁰. We identified CDK12 mutations, particularly in Exon 1, as the most frequent HRR-associated alteration in prostate cancer patients. However, the predictive value of this finding in the context of PARP-inhibitor therapy remains ambiguous; while the FDA has approved PARP inhibitors for CDK12-mutated prostate cancer, the EMA has not. A recent meta-analysis suggests a benefit of PARP inhibition in patients with CDK12 mutations, albeit less clear than in those with mutations in BRCA1/2. Moreover, different CDK12 mutations likely have variable impact and need to be studied in greater detail²¹.

Our data emphasizes the role of repeated sequential testing of cancer patients using liquid biopsy in the routine setting. In more than half of patients (56%), we identified a difference in biomarker profile between the first and a subsequent test. In the context of liquid biopsy, technical and pre-analytical considerations (e.g., variable concentrations of circulating tumor DNA) need to be taken into consideration alongside potential changes in tumor biology. Nevertheless, in approx. 80% of patients with differences between first and subsequent testing, we identified more or a higher tier of biomarkers in a subsequent test, indicating the potential clinical benefit of repeated testing.

Our data also highlights the potential of broad liquid-based NGS panels in identifying novel alterations, such as the ROS1::STX7 fusion identified in one patient with advanced prostate cancer. ROS1 fusions are detected in a number of malignancies, but have not yet been described in prostate cancer. Specific tyrosine kinase inhibitors are highly effective in malignancies with ROS1 fusions – however, since we were not able to validate our finding via RNA sequencing, its significance remains unclear.

In contrast to the analytical strengths of ctDNA testing, our findings highlight a central challenge in the clinical implementation of ctDNA profiling: the limited availability of structured clinical information to accompany molecular diagnostics. These findings should not be interpreted as a lack of clinical interest; rather, they reflect the reality of limited time, documentation burden, and the absence of structured fields in electronic health records to capture this information. In many cases, our ctDNA request forms lacked central clinical details—such as the cancer diagnosis, disease stage, prior systemic therapies, or the clinical indication for testing. In this regard, it is worth noting that the European Liquid Biopsy Society (ELBS) strongly recommends inclusion of clinical context, highlighting its importance for valid interpretation and integration of ctDNA into multidisciplinary care workflows²².

From the perspective of the diagnostic laboratory, contextual details such as current disease status, prior lines of therapy, suspected resistance mechanisms, and treatment goals are essential for meaningful interpretation of ctDNA results²². Without this information, even robust molecular findings may be misinterpreted, undervalued, or overlooked. Furthermore, the absence of clinical context can limit the lab's ability to provide interpretive reports that align with best practice guidelines or integrate with multidisciplinary care pathways, ultimately compromising the perceived reliability of the assay.

Given the high cost of liquid biopsy assays, rigorous justification of their clinical value is essential—particularly in publicly funded health systems. In our setting, the lack of standardized clinical data fields and MTB documentation made it difficult to determine whether and how ctDNA findings were acted upon or influenced patient care. This severely limits our ability to justify the cost of ctDNA testing to institutional stakeholders, insurers, and regulators, and may deter broader adoption despite the technical validity of the assay. This appears particularly unfortunate in light of our conclusion that repeated testing may be necessary to utilise the full diagnostic potential of ctDNA analysis, where testing costs may increase exponentially.

Accredited molecular diagnostics laboratories must demonstrate not only analytical validity but also appropriate clinical utility of tests within a defined use-case. When ctDNA results are returned without adequate clinical context, laboratories face challenges in meeting the interpretive and documentation requirements of regulatory and accrediting bodies (e.g., CLIA, CAP, ISO 15189).

Beyond immediate clinical use, ctDNA results hold significant value for retrospective biomarker analysis and real-world evidence generation. Yet in our study, the poor quality of accompanying clinical data undermined these opportunities, greatly reducing the research utility of both current and archived samples, and decreasing the return on investment for institutions and funders alike.

Importantly, the aim is not to shift blame to clinicians, but to acknowledge that current systems do not facilitate the seamless transfer of relevant clinical information. For diagnostics to fulfill their potential in precision oncology, institutional investment is needed in structured reporting, integrated ordering systems, and education on the importance of clinical annotation. These steps will not only improve diagnostic interpretation and patient outcomes, but also facilitate evidence generation, regulatory compliance, and cost-effectiveness assessments for emerging technologies such as ctDNA.

This study reaffirms that ctDNA analysis is analytically robust and identifies actionable genomic alterations in a significant proportion of patients. However, the poor quality of clinical documentation at our institution presents a major barrier to the assessment of clinical utility, cost justification, and research value of this promising technology. With improved clinical integration, ctDNA assays have the potential not only to guide individual care but also to generate large-scale data that can inform population-level cancer strategies.