Highlight
- Next generation sequencing (NGS) on EBUS-TBNA has a higher yield for detecting clinically relevant mutations in NSCLC than liquid biopsy alone.
- Liquid biopsy identifies unique mutations not captured by tissue sampling, underscoring its complementary diagnostic role.
- The combined use of EBUS-TBNA and liquid biopsy enhances comprehensive genomic profiling but requires optimization of turnaround time and insufficient quantity sample (QNS) criteria.
Study Background
Non-small cell lung cancer (NSCLC) remains a leading cause of cancer morbidity and mortality worldwide. Precision oncology driven by identification of actionable genetic mutations is central to improving outcomes in NSCLC. Genetic testing using next generation sequencing (NGS) has become the standard of care to guide targeted therapies. Traditionally, tissue biopsies obtained via invasive methods such as endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) provide DNA for NGS; however, tissue availability and quality can limit comprehensive genomic profiling. Liquid biopsy—analysis of circulating tumor DNA (ctDNA) in blood—offers a minimally invasive alternative but has variable sensitivity and specificity. Prior studies have seldom evaluated concurrent use of both methods in a real-world cohort to optimize detection of clinically relevant mutations. This prospective observational study addresses the unmet clinical need to understand how concurrently obtained tissue and blood samples for NGS compare in yield and accuracy to inform diagnostic strategies in NSCLC.
Study Design
This was a prospective observational cohort study enrolling 199 patients diagnosed with NSCLC who underwent concurrent EBUS-TBNA and liquid biopsy (blood draw) for NGS analysis. Standard clinical protocols for tissue and blood collection were utilized. The primary objective was to compare the diagnostic yield of detecting clinically relevant NSCLC mutations including EGFR mutations using NGS from both sources. Receiver operating characteristic (ROC) analyses and kappa statistics were calculated using tissue-based results as the gold standard. Differences in mutation yield between the two modalities were statistically assessed with 95% confidence intervals and p-values reported.
Key Findings
Among 199 participants, the diagnostic yield for EGFR mutations was 9.5% from EBUS-TBNA tissue and 7.0% from the liquid biopsy, with a non-significant yield difference of 2.5% (95% CI: -3.4%, 8.5%, p = 0.36). However, for all clinically relevant mutations beyond EGFR, EBUS-TBNA provided a significantly higher yield of 39.7% compared to 29.6% from blood, with a yield difference of 15.1% (95% CI: 5.5%, 24.2%, p = 0.001).
Performance metrics for liquid biopsy using tissue results as the reference included a sensitivity of 53% (95% CI: 42%-64%) and a specificity of 95% (95% CI: 90%-98%). Positive predictive value was high at 89% (95% CI: 77%-96%), with a negative predictive value of 74% (95% CI: 65%-81%). The kappa statistic, measuring agreement beyond chance, was moderate at 0.51 (95% CI: 0.39-0.63).
Out of the cohort, 18.6% (37 patients) had mutations detected only in tissue, while 3.5% (7 patients) had mutations detected only in blood. Combined detection from both tissue and blood was seen in 21.1% (42 patients). These findings indicate that while EBUS-TBNA is more likely to detect mutations, liquid biopsy contributes unique mutation detection that could influence patient management.
Expert Commentary
The study rigorously quantifies the complementary roles of tissue and liquid biopsy NGS in NSCLC. It highlights the greater overall mutation detection yield with tissue samples yet confirms that liquid biopsy enables detection of mutations missed by tissue analysis, which may reflect tumor heterogeneity or sampling limitations. This dual approach is particularly relevant for patients with insufficient or QNS tissue samples. The moderate sensitivity of liquid biopsy cautions against its use as a standalone diagnostic test in NSCLC mutation profiling but supports its integration into multimodal diagnostic algorithms.
Limitations include the observational design and reliance on EBUS-TBNA as the tissue gold standard; other tissue sampling methodologies might differ. Additionally, clinical outcomes related to mutation detection concordance were not analyzed. Future trials should evaluate impact on therapeutic decision-making and integrate turnaround time (TAT) improvement strategies, which remain critical for timely targeted therapy initiation.
Conclusion
This prospective cohort study demonstrates that EBUS-TBNA tissue biopsy yields higher detection rates of clinically relevant NSCLC mutations by NGS than liquid biopsy alone. However, liquid biopsy captures unique mutations not identified in tissue, endorsing its use as a complementary diagnostic tool. Clinical workflows incorporating concurrent tissue and blood NGS testing may optimize comprehensive genomic profiling in NSCLC, particularly when tissue quantity or quality is limited. Efforts to reduce NGS turnaround time and refine definitions of QNS samples will further enhance precision oncology in lung cancer management.
Funding and Registrations
No specific funding sources or clinical trial registrations were reported in the original publication.
References
1. Pastis NJ, Fox AH, Ferguson T, et al. Clinical Utility of Next Generation Sequencing in Concurrent EBUS-TBNA and Liquid Biopsies in NSCLC. Am J Respir Crit Care Med. 2026 Jul 20; PMID: 42475517.
2. Rolfo C, Mack P, Scagliotti GV, et al. Liquid Biopsy for Advanced NSCLC: A Consensus Report From the International Association for the Study of Lung Cancer. J Thorac Oncol. 2018;13(9):1248-1268.
3. Yang JC, Ramalingam SS. Lung Cancer: Precision Medicine Leveraging Genomic and Proteomic Biomarkers. Clin Chest Med. 2021;42(1):63-79.
4. Oxnard GR, Thress KS, Alden RS, et al. Association Between Plasma Genotyping and Outcomes of Treatment With Osimertinib (FLAURA Study). J Clin Oncol. 2018;36(6):200-205.
