Hayoung Seong; Soo Han Kim; Mi-Hyun Kim; Ahrong Kim; Ju Sun Song; Jung Seop Eom

doi:10.4143/crt.2025.114

Articles

Page Path: HOME > Cancer Res Treat > Volume 58(3); 2026 > Article

Original Article Lung and Thoracic cancer Genomic Insights into Innate Resistance in Early-Stage EGFR-Mutant Non–Small Cell Lung Cancer: A Comprehensive Analysis of Next-Generation Sequencing Data: Hayoung Seong^1,2,3, Soo Han Kim^1,2, Mi-Hyun Kim^1,2, Ahrong Kim⁴, Ju Sun Song⁵, Jung Seop Eom^1,2,3; Cancer Research and Treatment : Official Journal of Korean Cancer Association 2026;58(3):738-747.
DOI: https://doi.org/10.4143/crt.2025.114
Published online: August 25, 2025

¹Department of Internal Medicine, Pusan National University School of Medicine, Busan, Korea

²Department of Internal Medicine, Pusan National University Hospital, Busan, Korea

³Biomedical Research Institute, Pusan National University Hospital, Busan, Korea

⁴Department of Pathology, Pusan National University Hospital, Busan, Korea

⁵GC Genome Corporation, Yongin, Korea

Correspondence: Jung Seop Eom, Department of Internal Medicine, Pusan National University School of Medicine, 179 Gudeok-ro, Seo-gu, Busan 49241, Korea
Tel: 82-51-240-7889 E-mail: ejspulm@gmail.com

• Received: January 28, 2025 • Accepted: August 19, 2025

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

2,219 Views
113 Download
1 Scopus

prev next

Full Article

Download PDF

Abstract

Purpose
Comprehensive genomic profiling of early-stage non–small cell lung cancer (NSCLC) with epidermal growth factor receptor (EGFR) mutations remains limited. This study aimed to investigate genomic profiles of early- and advanced-stage EGFR-mutant NSCLC and identify potential innate resistance mechanisms to EGFR-tyrosine kinase inhibitors (TKIs) using targeted next-generation sequencing (NGS).
Materials and Methods
This retrospective observational study analyzed genomic profiles of patients with early-stage (IA-IIIA) and advanced-stage (IIIB-IV) EGFR-mutant NSCLC from the Lung Cancer NGS registry. Targeted NGS was performed to assess concurrent genetic alterations (GAs), tumor mutational burden (TMB), and variant allele frequency (VAF) of EGFR mutations.
Results
Overall, 160 patients (100 early-stage and 60 advanced-stage) were analyzed. The proportion of patients with concurrent GAs was not significantly different between stages (82.0% vs. 91.7%, p=0.092). Median TMB was 3.8 mutations/Mb in both stages, with no significant difference (p=0.206). However, the median VAF of EGFR mutations was significantly lower in early-stage compared to that in advanced-stage (19.3% vs. 29.6%, p=0.002). While TMB remained unchanged with disease progression (p=0.192), VAF of EGFR mutations increased significantly (p < 0.001). Moreover, the frequencies of concurrent single nucleotide variants and copy number variants were significantly lower in early-stage NSCLC.
Conclusion
Genomic heterogeneity in EGFR-mutant NSCLC arises early in tumorigenesis. The comparable TMB and lower VAF of EGFR mutations in early-stage disease suggest that innate resistance to EGFR-TKIs may be driven by concurrent GAs, supporting the consideration of combination therapies even in early-stage EGFR-mutant NSCLC.
Key words: Lung neoplasms, EGFR mutation, Tyrosine kinase inhibitors, High-throughput nucleotide sequencing, Genetic profile

Introduction

Lung cancer remains one of the most prevalent and deadly malignancies worldwide [1]. Despite advances in surgical treatments for early-stage non–small cell lung cancer (NSCLC), 30%-55% of patients experience recurrence or succumb to death, presenting a significant obstacle to achieving long-term survival [2,3]. Although adjuvant cytotoxic chemotherapy has been considered the standard post-surgical treatment, its impact remains limited, with only a 5.4% absolute 5-year survival benefit, as shown in the LACE trial [4]. Additionally, postoperative radiotherapy has not demonstrated a survival benefit in resected NSCLC [5].

Recently, adjuvant osimertinib has shown significant improvements in both disease-free and overall survival compared with placebo in patients with resected NSCLC harboring common epidermal growth factor receptor (EGFR) mutations [6,7]. However, uncertainty remains regarding which patients may benefit from adjuvant cytotoxic chemotherapy prior to adjuvant osimertinib treatment. The ongoing phase III NeoADAURA trial is addressing this question by including three arms: neoadjuvant osimertinib, neoadjuvant osimertinib combined with chemotherapy, and neoadjuvant chemotherapy, exploring the need for a combined EGFR–tyrosine kinase inhibitor (EGFR-TKI) and chemotherapy approach [8].

It remains largely unknown which patients are sufficiently treated with EGFR-TKI monotherapy and which may require combination strategies. Concurrent genetic alterations (GAs), such as TP53 mutations or MET amplifications, have been suggested as potential causes of innate resistance to EGFR-TKI [9,10]. These alterations may serve as biomarkers to guide the use of combination treatments, such as the osimertinib and chemotherapy combination in the FLAURA2 trial or the lazertinib and amivantamab combination in the MARIPOSA trial [11,12]. Recent studies have demonstrated that concurrent GAs are frequently present even in early-stage EGFR-mutant NSCLC, with reported rates ranging from 86.9% to 94.1%, and are associated with shorter disease-free survival [13,14]. However, comprehensive genomic profiling studies that directly compare early- and advanced-stage EGFR-mutant NSCLC remain scarce. Therefore, this study aimed to utilize next-generation sequencing (NGS) to compare the genomic profiles of patients with early- and advanced-stage EGFR-mutant NSCLC and to investigate potential mechanisms of innate resistance to EGFR-TKI in early-stage disease.

Materials and Methods

1. Study design and patient selection

This study was conducted at a 1,200-bed university-affiliated tertiary care hospital in Busan, Korea, between January 1, 2022, and October 31, 2023. All patients diagnosed with NSCLC who underwent NGS analysis were prospectively registered in the Pusan National University Hospital Lung Cancer NGS Registry. Data were extracted from the Lung Cancer NGS Registry and retrospectively analyzed on June 1, 2024. The inclusion criteria focused on patients with EGFR-mutant NSCLC who had undergone targeted NGS analysis using tissue samples. Patients were excluded if they had undergone NGS testing owing to treatment failure or had insufficient tissue for analysis.

2. Targeted gene NGS and interpretation

DNA was isolated from formalin-fixed, paraffin-embedded tissue sections with a strict tumor purity threshold of over 30%, using the QIAamp DNA Kit (QIAGEN GmbH). The double-stranded DNA (dsDNA) strand was quantified using the Qubit Fluorometer with the Qubit dsDNA HS Assay Kit (Thermo Fisher Scientific). Library preparation was performed using the multiplex polymerase chain reaction (PCR)–based Ion Torrent AmpliSeq technology (Life Technologies, Thermo Fisher Scientific) with the Oncomine Comprehensive Assay Plus (Ion Torrent, Thermo Fisher Scientific), which covers 501 cancer-associated genes. This assay detects single nucleotide variants (SNVs), small insertions/deletions (Indels), amplifications, fusions, immunotherapeutic markers, tumor mutational burden (TMB), and microsatellite instability. For RNA library preparation, 20 ng of RNA was mixed with two primer pools and AmpliSeq HiFi Master Mix, then transferred to a PCR cycler (SimpliAmp Thermal Cycler, Life Technologies, Thermo Fisher Scientific). After PCR, RNA primer end sequences were partially digested with the FuPa reagent, followed by ligation of barcoded sequencing adapters (Ion Xpress Barcode Adapters, Life Technologies, Thermo Fisher Scientific). The final libraries were purified using Celemics MagBeads (Celemics) and quantified via quantitative polymerase chain reaction. For DNA library preparation, a deamination reaction was applied in the OCA-Plus assay (Thermo Fisher Scientific) on extracted DNA using heat-labile uracil-DNA glycosylase before PCR amplification. Sequencing was performed using the Ion Torrent S5 XL platform (Thermo Fisher Scientific) with the Ion 550 Chip Kit (Thermo Fisher Scientific) following the manufacturer’s instructions.

Genomic data were analyzed using Ion Reporter Software v5.6 (Thermo Fisher Scientific) to detect SNVs, Indels, amplifications, and potential fusions. Quality assessments were based on coverage analysis reports from the Ion Reporter Software, including mapped reads, mean depth, uniformity, and alignment over the target region. Adequate quality control metrics were defined as a mean depth of coverage of 1,200× and a minimum depth of 250×, with a limit of detection of 5% variant allele frequency (VAF), as determined on analytical performance validation testing. For RNA fusion detection and copy number alterations, quality control criteria included total mapped fusion panel reads ≥ 500,000, with pool-1 and pool-2 each having ≥ 100,000 mapped fusion reads, and a mean absolute pairwise difference score < 0.5, per the manufacturer’s guidelines. Amplification was defined as an average copy number ≥ 4. Identified mutations were manually reviewed using the Golden Helix GenomeBrowse visualization tool ver. 3.0.0 (Golden Helix, Inc., https://www.goldenhelix.com). Mutations detected in the test were not reconfirmed using other methods, such as Sanger sequencing.

Variant interpretation was performed according to the guidelines from the Association for Molecular Pathology, American Society of Clinical Oncology, and College of American Pathologists [15]. VAF was calculated as the proportion of sequence reads supporting a specific DNA variant relative to the total number of reads covering that position. TMB was calculated as the number of somatic, non-synonymous, coding base substitutions, and indels per megabase of genome analyzed, using the Ion Reporter Software v5.6 (Thermo Fisher Scientific). The Oncomine Comprehensive Assay Plus panel covers approximately 1.2 Mb of coding regions relevant to TMB estimation. Synonymous mutations and known germline variants were excluded from the calculation.

3. Concurrent GA and signaling pathway analysis

The presence and number of concurrent GAs beyond EGFR mutations, as well as the number of affected signaling pathway alterations (SPAs), were analyzed. These pathways, identified from The Cancer Genome Atlas [16], included (1) cell cycle, (2) Hippo signaling, (3) Myc signaling, (4) Notch signaling, (5) oxidative stress response/NRF2, (6) phosphoinositide 3-kinase (PI3K)/Akt signaling, (7) receptor tyrosine kinase–RAS/mitogen-activated protein kinase signaling, (8) transforming growth factor β (TGFβ) signaling, (9) p53, and (10) β-catenin/Wnt signaling. The number of SPAs was categorized as ‘low’ if one or two pathways were affected and ‘high’ if three or more pathways were involved [17,18].

4. Data collection

Demographic and clinical data, including age, sex, histopathology, TNM stage (according to the eighth edition of the American Joint Committee on Cancer TNM staging), and sample collection method, were gathered from patients’ electronic medical records. Results from the NGS analysis were sourced from the Lung Cancer NGS registry. Stages IA to IIIA were classified as early-stage NSCLC, and stages IIIB to IV as advanced-stage NSCLC. EGFR mutation subtypes were categorized as (1) common EGFR mutations (EGFR exon 19 deletion or EGFR exon 21 L858R point mutation), (2) EGFR exon 20 insertion mutations, and (3) uncommon EGFR mutations. When both common and uncommon mutations were present, they were classified as common EGFR mutations.

5. Outcomes

The primary outcome was to compare the overall genomic profiles between early- and advanced-stage EGFR-mutant NSCLC. Secondary outcomes included: (1) TMB; (2) VAF of EGFR mutations; and (3) presence of concurrent SNVs, copy number variations (CNVs), and Indels. Subgroup analyses were additionally conducted to explore factors associated with postoperative recurrence or death in early-stage EGFR-mutant NSCLC and factors associated with the overall response rate to EGFR-TKIs treatment in advanced-stage EGFR-mutant NSCLC. In addition to the primary comparison between early- and advanced-stage groups, a supplementary subgroup analysis was conducted to compare genomic features among stage I, stage II-III, and stage IV diseases, based on their distinct prognostic and therapeutic profiles (Supplementary Material).

6. Statistical analysis

Continuous variables are reported as median (interquartile range) and were compared using Student’s t test or the Mann-Whitney U test, as appropriate. The Kruskal-Wallis test was used to assess differences in continuous variables across stages. When statistical significance was detected, the Jonckheere-Terpstra test was applied for post hoc analysis. Categorical variables are reported as number (%) and were compared using the chi-square or Fisher’s exact test, as appropriate. Univariate and multivariable Cox proportional hazards regression analyses were performed to identify risk factors associated with recurrence or mortality. Multivariate logistic regression analysis was conducted to determine factors associated with the overall response to EGFR-TKIs. All tests were two-tailed, and p < 0.05 indicated statistical significance. Statistical analyses were conducted using the R statistical software ver. 4.3.3 (R Core Team, 2024) and additional packages (maftools).

Results

1. Study participants

During the study period, 643 patients were registered in the Lung Cancer NGS registry. Of these, 160 patients with EGFR-mutant NSCLC were included in the study, consisting of 100 patients with early-stage NSCLC and 60 with advanced-stage NSCLC (S1 Fig.). Common EGFR mutations were detected in 90.6% (145/160) of the patients, with three of them having concurrent uncommon EGFR mutations, all of which were de novo EGFR T790M mutations. Exon 20 insertions were observed in 6.9% (11/160) of patients, and uncommon EGFR mutations in 2.5% (4/160) (S2 Fig.). Table 1 summarizes the baseline characteristics of the 160 patients with EGFR-mutant NSCLC. Baseline characteristics did not differ between early- and advanced-stage patients, except for the method of specimen collection for NGS analysis.

2. Overall genomic profile

The top 30 concurrent GAs are displayed in S3 Fig. Among patients analyzed using targeted NGS, 82.0% (82/100) of early-stage and 91.7% (55/60) of advanced-stage patients had concurrent GAs other than EGFR mutations, with no significant difference (p=0.092). However, the median number of concurrent GAs and proportion of patients with high SPAs (≥ 3) were significantly lower in early-stage than in advanced-stage NSCLC (median GAs: 2 vs. 3, p=0.002; high SPAs: 17.0% vs. 40.0%, p=0.001) (Table 2, Fig. 1). When comparing the frequencies of concurrent SPAs, no significant differences were found in the p53, Hippo, TGFβ, and Notch pathways between early- and advanced-stage NSCLC. However, significantly lower frequencies of SPAs were observed in the cell cycle, PI3K, Myc, and Wnt pathways in early- than in advanced-stage NSCLC.

3. TMB

The median TMB was 3.8 mutations/Mb in both early- and advanced-stage NSCLC, with no significant difference between the groups (p=0.206). This lack of variation in TMB was consistent across EGFR mutation subtypes (Table 3). Additionally, this finding indicates that disease progression does not significantly alter the median TMB (Fig. 2).

4. VAF of EGFR mutation

The median VAF of EGFR mutations was significantly lower in early- than in advanced-stage NSCLC (19.3% vs. 29.6%, p=0.002) (Table 4). Specifically, in patients with common EGFR mutations, the median VAF was notably lower in early- than in advanced-stage NSCLC (19.0% vs. 29.9%, p< 0.001). Moreover, the median VAF increased with disease progression (Fig. 2).

5. Concurrent SNVs

The proportion of patients with concurrent SNVs other than EGFR mutations was lower in early- than in advanced-stage NSCLC (58.0% vs. 80.0%, p=0.004), with a particularly lower proportion of missense mutations in early-stage cases (39.0% vs. 63.3%, p=0.003). S4 Table lists oncogenic GAs that occurred in at least 5% of cases, with TP53 being the most common oncogenic SNV, followed by RBM10. Importantly, no significant differences were observed between early- and advanced-stage NSCLC in the frequencies of concurrent SNVs of TP53 and RBM10 (31.0% vs. 41.7%, p=0.171 for TP53; 6.0% vs. 3.3%, p=0.711 for RBM10). The VAF of oncogenic SNVs did not differ by stage, as shown in the detailed distribution of concurrent SNVs in S5A-S5D Fig.

6. Concurrent CNVs

The proportion of patients with concurrent CNVs was lower in early- than in advanced-stage NSCLC (31.0% vs. 51.7%, p=0.009). EGFR amplification was the most frequent oncogenic CNV, followed by amplifications in RAC1, FOXA1, TERT, and MDM2 (S4 Table). Notably, the proportions of patients with TERT, MDM2, and DDR2 amplifications did not significantly differ between early- and advanced-stage NSCLC. However, amplifications of EGFR, RAC1, FOXA1, and MET were less frequent in early- than in advanced-stage NSCLC. The copy number gain of oncogenic CNVs did not vary by stage, as illustrated in the detailed distribution of concurrent CNVs in S5E-S5R Fig.

7. Concurrent Indels

No significant difference was found in the proportion of patients with concurrent Indels between early- and advanced-stage NSCLC (26.0% vs. 31.7%, p=0.440). Additionally, no significant differences were observed in the distribution of Indel subtypes between stages. RBM10 was the most common oncogenic Indel (S4 Table). No significant differences were found between early- and advanced-stage NSCLC in the frequency of concurrent RBM10 Indels (8.0% vs. 6.7%, p > 0.99). The detailed distribution of concurrent Indels is shown in S5S-S5T Fig.

8. Subgroup analyses

Overall, 98 patients with early-stage EGFR-mutant NSCLC underwent complete surgical resection. The median postoperative follow-up duration was 25.8 months (range, 4.3 to 36.0). Among these patients, 31 received adjuvant treatment, including chemotherapy alone (n=16), EGFR-TKI alone (n=11), chemotherapy followed by EGFR-TKI (n=3), or concurrent chemoradiotherapy (n=1). During the follow-up period, 17 patients experienced recurrence or death. In the multivariate analysis, disease stage remained the only independent factor significantly associated with an increased risk of recurrence or death (S6 Table).

Among the 60 patients with advanced-stage EGFR-mutant NSCLC, 43 received EGFR-TKI therapy and were evaluable for response. The remaining patients included seven with stage IIIB disease who underwent surgery as initial treatment, five with exon 20 insertions who received chemotherapy, three who were untreated due to poor general condition, and two who were lost to follow-up. The median follow-up duration was 20.5 (range, 2.1 to 32.0) months. During this period, the median progression-free survival was 18.7 months (95% confidence interval, 13.7 to 24.0). The objective response rate was 58.1% (25/43), and the disease control rate was 90.7% (39/43). In multivariate logistic regression analysis, male sex, presence of concurrent TP53 SNVs, and lower VAF of EGFR mutations were independently associated with an increased risk of progression or death (R²=0.335, p < 0.001). Furthermore, male sex and lower VAF of EGFR mutations were significantly associated with poorer response to EGFR-TKI (R²=0.272, p=0.007) (S7 Table, S8 Fig.).

Discussion

In this study, we provide a comprehensive genomic profile of early-stage EGFR-mutant NSCLC using targeted NGS analysis. Our findings indicated that 82.0% of patients with early-stage EGFR-mutant NSCLC exhibited concurrent GAs, a proportion not significantly different from the 91.7% observed in advanced-stage patients. Although no significant difference was observed in the TMB between early- and advanced-stage, the VAF of EGFR mutations increased as the disease progressed. In contrast, no statistically significant differences were observed in the VAF or copy number gains of oncogenic concurrent GAs between early- and advanced-stage disease. This suggests that genomic heterogeneity in EGFR-mutant NSCLC emerges early in tumorigenesis, and alterations beyond EGFR mutations may influence the response to EGFR-TKI treatment in early-stage disease.

As demonstrated in the ADAURA trial, even early-stage EGFR-mutant NSCLC patients treated with adjuvant osimertinib after surgical resection showed a 3-year disease-free survival rate of 84%, highlighting a 16% risk of recurrence or death, possibly owing to innate resistance to EGFR-TKIs [6]. In advanced-stage EGFR-mutant NSCLC, concurrent GAs and SPAs are associated with poor clinical outcomes [19,20]. However, research exploring the relationship between clonal heterogeneity, innate EGFR-TKI resistance, and survival outcomes in early-stage patients remains limited.

Previous studies have identified the presence of concurrent GAs in early-stage EGFR-mutant NSCLC. For instance, Zhao et al. [13], using NGS targeting 416 genes, detected concurrent GAs in 94.1% of early-stage patients, with a median of 3.7 alterations, associating these with a shorter disease-free survival. Similarly, Deng et al. [14], using NGS targeting 285 genes, reported concurrent GAs in 86.9% of stage IIB-IIIA patients, with a median of two alterations, suggesting that the number of GAs might influence survival outcomes. In our study, using NGS targeting 501 genes, 82.0% of early-stage EGFR-mutant NSCLC patients exhibited concurrent GAs, with a median of two alterations. To our knowledge, this is the first study to directly compare genomic profiles of early- and advanced-stage EGFR-mutant NSCLC, showing no significant difference in the prevalence of concurrent GAs. These findings reinforce the concept of substantial genomic heterogeneity from the onset of EGFR-mutant NSCLC, potentially contributing to innate EGFR-TKI resistance.

TMB measures the total number of somatic mutations per megabase, while VAF reflects the proportion of DNA sequencing reads containing a specific GA. Our results show that although the median TMB remained constant as the disease progressed, the median VAF of EGFR mutations increased significantly with advanced stages (Fig. 2). This pattern suggests progressive clonal expansion of EGFR-mutant cells during tumor advancement, despite an unchanged overall mutation rate per megabase. Concurrently, there was selective enrichment of specific subclonal copy number amplifications, notably RAC1 and FOXA1, in advanced-stage disease. However, the individual VAFs and copy number gains of these alterations did not significantly differ from those in early-stage disease, indicating their predominantly subclonal nature. Historical studies suggest that a higher abundance of EGFR mutations may predict a better response to EGFR-TKIs [21,22]. Thus, the higher burden of non-EGFR mutations in early-stage disease may contribute to a reduced response to these inhibitors. Supporting this notion, a phase II trial of neoadjuvant osimertinib in patients with resectable stage I-IIIA NSCLC showed a lower overall response of 52%, compared with the 80% observed in the FLAURA trial for advanced EGFR-mutant NSCLC [23,24]. This implies that the lower EGFR mutational burden in early-stage disease might be associated with a higher likelihood of innate resistance to EGFR-TKIs. Moreover, the frequencies, VAFs, or copy number gains of oncogenic GAs, such as TP53 and RBM10 mutations, and amplifications of MDM2, TERT, and DDR2, did not significantly differ across stages. Their stable presence suggests early establishment and maintenance of genomic heterogeneity. Given the complexity and persistence of such concurrent GAs, early-stage EGFR-mutant NSCLC may benefit from combination therapies that target these diverse resistance mechanisms [25-28], potentially enhancing treatment effectiveness and overcoming innate resistance.

This study has some limitations. First, this study was based on a retrospective observational design without a predefined control group. Although additional subgroup analyses were performed to explore clinical factors associated with outcomes in early- and advanced-stage EGFR-mutant NSCLC, the absence of a prospectively-matched control cohort and the lack of propensity score matching may have introduced selection bias and limited the ability to establish causal relationships. Second, differences in tumor purity between surgical specimens and small biopsies could influence NGS results; however, this was mitigated by a stringent tumor cellularity threshold of > 30% for all included samples and excluded those that failed quality control. Nevertheless, differences in specimen collection methods may have influenced VAF measurements. Variations in tumor cell content, DNA quality, and technical processing could affect the accuracy and consistency of VAF results. Although we observed the consistent trend of increasing VAF with disease progression across both specimen types, the potential impact of sample variability remains a limitation of our study. Future studies directly comparing different specimen types obtained from the same patients are warranted to better elucidate the effect of sampling methods on VAF analysis. Third, PCR amplification-based target enrichment methods, while widely used for targeted NGS panels in clinical oncology, have intrinsic limitations compared to hybrid capture-based methods. These include lower uniformity of sequencing coverage and a higher risk of false-positive or false-negative variant calls [29]. Nevertheless, recent studies have demonstrated that PCR amplification-based targeted NGS panels can achieve excellent sensitivity, specificity, and robustness even in clinical samples with low DNA quantity and quality [30]. In this study, PCR amplification-based targeted NGS was selected as it was the only clinically validated method available at our institution and had demonstrated high sequencing depth and sufficient performance for detecting SNVs, Indels, and CNVs in formalin-fixed paraffin-embedded specimens. To ensure the reliability and robustness of variant detection, stringent quality control criteria were applied, including minimum sequencing depth thresholds and strict variant calling standards. Fourth, our findings are based on data from a single center, which may limit their generalizability. Additionally, while comparing survival outcomes based on genetic variants is important, our study was limited by a short follow-up period, requiring further analysis.

This study highlights the significant genomic heterogeneity present in EGFR-mutant NSCLC from early tumor development. Our findings suggest that concurrent GAs and SPAs may play critical roles in innate resistance to EGFR-TKIs in early-stage disease. Despite similar TMB scores between early- and advanced-stage patients, the lower VAF of EGFR mutations in early-stage disease suggests that combination treatment strategies could be a valuable option to overcome innate resistance. These results emphasize the importance of comprehensive genomic profiling to inform personalized treatment strategies for early-stage EGFR-mutant NSCLC.

Electronic Supplementary Material

Supplementary materials are available at Cancer Research and Treatment website (https://www.e-crt.org).

crt-2025-114_S1_Fig.pdf

crt-2025-114_S2_Fig.pdf

crt-2025-114_S3_Fig.pdf

crt-2025-114_S4_Table.pdf

crt-2025-114_S5_Fig.pdf

crt-2025-114_S6_Table.pdf

crt-2025-114_S7_Table.pdf

crt-2025-114_S8_Fig.pdf

crt-25-114-Supplementary-Material.pdf

NOTES

Ethical Statement

Approval for the study was obtained from the Institutional Review Board of Pusan National University Hospital (IRB No. 2409-005-142), with the patient consent requirement waived owing to the retrospective and observational nature of the study. The study adhered to the ethical principles outlined in the Declaration of Helsinki (1975) and followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.

Author Contributions

Conceived and designed the analysis: Seong H, Kim SH, Kim MH, Eom JS.

Collected the data: Seong H, Kim A, Song JS, Eom JS.

Contributed data or analysis tools: Seong H, Kim SH, Kim MH.

Performed the analysis: Seong H, Kim SH, Kim MH, Kim A, Song JS, Eom JS.

Wrote the paper: Seong H, Eom JS.

Conflicts of Interest

Conflict of interest relevant to this article was not reported.

Funding

This research was supported by a clinical research grant from the Pusan National University Hospital 2025.

Acknowledgments

We would like to thank Editage (www.editage.co.kr) for editing and reviewing this manuscript for English language.

Fig. 1.

Signaling pathway alterations in the overall study population. The figure illustrates the frequency and distribution of genomic alterations classified by canonical oncogenic signaling pathways, including RTK-RAS/MAP kinase, p53, cell cycle, PI3K, Myc, Wnt, Hippo, TGFβ, and Notch. Pathway classification was based on The Cancer Genome Atlas framework. Alterations were identified using targeted next-generation sequencing. MAP, mitogen-activated protein; PI3K, phosphoinositide 3-kinase; RTK, receptor tyrosine kinase; SPA, signaling pathway alteration; TGFβ, transforming growth factor β.

Fig. 2.

Tumor mutational burden and variant allele frequency (VAF) of epidermal growth factor receptor (EGFR) mutations in non–small cell lung cancer (NSCLC). The figure compares the tumor mutational burden (TMB) and VAF of EGFR mutations between early- and advanced-stage NSCLC. TMB is expressed as mutations per megabase and VAF represents the proportion of sequencing reads supporting the EGFR mutation.

Table 1.

Baseline characteristics of EGFR-mutant NSCLC

Value	Total (n=160)	Early stage (n=100)	Advanced stage (n=60)	p-value
Age at diagnosis (yr)	68.0 (61-75)	67.0 (59-73)	69.5 (63-75)	0.112
Sex
Female	102 (63.8)	61 (61.0)	41 (68.3)	0.350
Male	58 (36.2)	39 (39.0)	19 (31.7)
Stage
IA	36 (22.5)	36 (36.0)	-	-
IB	25 (15.6)	25 (25.0)	-
IIA	15 (9.4)	15 (15.0)	-
IIB	15 (9.4)	15 (15.0)	-
IIIA	9 (5.6)	9 (9.0)	-
IIIB	7 (4.4)	-	7 (11.7)
IIIC	0	-	0
IV	53 (33.1)	-	53 (88.3)
Specimen collection
Surgical specimen	108 (67.5)	94 (94.0)	14 (23.3)	< 0.001
Small biopsy	52 (32.5)	6 (6.0)	46 (76.7)
Histopathology
Adenocarcinoma	154 (96.3)	97 (97.0)	57 (95.0)	0.197
Adenosquamous cell carcinoma	2 (1.3)	2 (2.0)	0
Squamous cell carcinoma	2 (1.3)	1 (1.0)	1 (1.7)
NSCLC NOS	2 (1.3)	0	2 (3.3)
EGFR mutation
Common EGFR mutation^a)	145 (90.6)	91 (91.0)	54 (90.0)	0.753
EGFR exon 20 insertion	11 (6.9)	6 (6.0)	5 (8.3)
Uncommon EGFR mutation	4 (2.5)	3 (3.0)^b)	1 (1.6)^c)
Smoking status
Never smoker	116 (72.5)	71 (71.0)	45 (75.0)	0.583
Ever smoker	44 (27.5)	29 (29.0)	15 (25.0)
Predominant pattern^d)
Lepidic	5 (4.5)	5 (5.1)	0	0.653
Acinar or papillary	99 (90.0)	89 (89.8)	10 (90.9)
Micropapillary	6 (5.5)	5 (5.1)	1 (9.1)
Differentiation^e)
Well-differentiated	15 (13.5)	15 (15.2)	0	< 0.001
Moderately differentiated	76 (68.5)	71 (71.7)	5 (41.7)
Poorly differentiated	20 (18.0)	13 (13.1)	7 (58.3)

Values are presented as median (interquartile range) or number (%). EGFR, epidermal growth factor receptor; NSCLC NOS, non-small cell lung cancer not otherwise specified.

^a) Concomitant de novo EGFR T790M mutations were found in two patients with early-stage NSCLC and one patient with advanced-stage NSCLC,

^b) One patient with L861Q, one with S768I and G724S, and one with G719A,

^c) One patient with G719A,

^d) Predominant pattern could not be identified in one patient in the early-stage and 49 patients in the advanced-stage,

^e) Differentiation could not be identified in one patient in the early-stage and 48 patients in the advanced-stage.

Table 2.

Overall genomic profile of EGFR-mutant NSCLC

Value	Early stage (n=100)	Advanced stage (n=60)	p-value
No. of patients with concurrent GAs	82 (82.0)	55 (91.7)	0.092
Median No. of concurrent GAs	2 (1-4)	3 (2-9)	0.002
Concurrent SPAs
Low (1-2)	83 (83.0)	36 (60.0)	0.001
High (≥ 3)	17 (17.0)	24 (40.0)
Type of SPA
RTK/RAS/MAP-kinase	100 (100)	60 (100)	N / A
p53	42 (42.0)	33 (55.0)	0.065
Cell cycle	12 (12.0)	19 (31.7)	0.002
PI3K	11 (11.0)	17 (28.3)	0.005
Myc	3 (3.0)	9 (15.0)	0.005
Wnt	3 (3.0)	7 (11.7)	0.041
Hippo	0	3 (5.0)	0.051
TGFβ	0	3 (5.0)	0.051
Notch	1 (1.0)	0	> 0.99
NRF2	0	0	N / A

Values are presented as number (%) or median (interquartile range). EGFR, epidermal growth factor receptor; GAs, genetic alterations; MAP, mitogen-activated protein; N/A, not applicable; NSCLC, non–small cell lung cancer; PI3K, phosphoinositide 3-kinase; RTK, receptor tyrosine kinase; SPAs, signaling pathway alterations; TGFβ, transforming growth factor β.

Table 3.

TMB score

EGFR mutation subtypes	Early stage (n=100)	Advanced stage (n=60)	p-value
Total subjects	3.8 (2.9-4.8)	3.8 (2.8-5.7)	0.206
Common EGFR mutation	3.8 (2.9-4.8)	4.3 (2.8-5.7)	0.141
EGFR exon 20 insertion	3.3 (2.1-5.9)	2.9 (1.0-2.9)	0.292
Uncommon EGFR mutation	4.8 (3.8-5.7)	7.6 (7.6-7.6)	0.327

Values of the tumor mutational burden (TMB) score are presented as mutations per megabase, with the median (interquartile range). EGFR, epidermal growth factor receptor.

Table 4.

VAF of EGFR mutation

EGFR mutation subtypes	Early stage (n=100)	Advanced stage (n=60)	p-value
Total subjects	19.3 (12.5-29.4)	29.6 (18.8-44.8)	0.002
Common EGFR mutation	19.0 (12.5-29.7)	29.9 (19.1-44.9)	< 0.001
EGFR exon 20 insertion	25.7 (16.4-29.4)	14.7 (9.9-21.3)	0.403
Uncommon EGFR mutation	16.8 (13.2-20.6)	32.7 (32.7-32.7)	0.206

Values of the variant allele frequency (VAF) are presented as percentages, with the median (interquartile range). EGFR, epidermal growth factor receptor.

REFERENCES

1. Bray F, Laversanne M, Sung H, Ferlay J, Siegel RL, Soerjomataram I, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2024;74:229–63. Article PubMed PMC
2. Hoffman PC, Mauer AM, Vokes EE. Lung cancer. Lancet. 2000;355:479–85. Article PubMed
3. Rajaram R, Huang Q, Li RZ, Chandran U, Zhang Y, Amos TB, et al. Recurrence-free survival in patients with surgically resected non-small cell lung cancer: a systematic literature review and meta-analysis. Chest. 2024;165:1260–70. Article PubMed
4. Pignon JP, Tribodet H, Scagliotti GV, Douillard JY, Shepherd FA, Stephens RJ, et al. Lung adjuvant cisplatin evaluation: a pooled analysis by the LACE Collaborative Group. J Clin Oncol. 2008;26:3552–9. Article PubMed
5. Le Pechoux C, Pourel N, Barlesi F, Lerouge D, Antoni D, Lamezec B, et al. Postoperative radiotherapy versus no postoperative radiotherapy in patients with completely resected non-small-cell lung cancer and proven mediastinal N2 involvement (Lung ART): an open-label, randomised, phase 3 trial. Lancet Oncol. 2022;23:104–14. PubMed
6. Wu YL, Tsuboi M, He J, John T, Grohe C, Majem M, et al. Osimertinib in resected EGFR-mutated non-small-cell lung cancer. N Engl J Med. 2020;383:1711–23. PubMed
7. Tsuboi M, Herbst RS, John T, Kato T, Majem M, Grohe C, et al. Overall survival with osimertinib in resected EGFR-mutated NSCLC. N Engl J Med. 2023;389:137–47. PubMed
8. Tsuboi M, Weder W, Escriu C, Blakely C, He J, Dacic S, et al. Neoadjuvant osimertinib with/without chemotherapy versus chemotherapy alone for EGFR-mutated resectable non-small-cell lung cancer: NeoADAURA. Future Oncol. 2021;17:4045–55. Article PubMed PMC
9. Vokes NI, Chambers E, Nguyen T, Coolidge A, Lydon CA, Le X, et al. Concurrent TP53 mutations facilitate resistance evolution in EGFR-mutant lung adenocarcinoma. J Thorac Oncol. 2022;17:779–92. Article PubMed PMC
10. Coleman N, Hong L, Zhang J, Heymach J, Hong D, Le X. Beyond epidermal growth factor receptor: MET amplification as a general resistance driver to targeted therapy in oncogene-driven non-small-cell lung cancer. ESMO Open. 2021;6:100319.Article PubMed PMC
11. Planchard D, Janne PA, Cheng Y, Yang JC, Yanagitani N, Kim SW, et al. Osimertinib with or without chemotherapy in EGFR-mutated advanced NSCLC. N Engl J Med. 2023;389:1935–48. Article PubMed PMC
12. Cho BC, Lu S, Felip E, Spira AI, Girard N, Lee JS, et al. Amivantamab plus lazertinib in previously untreated EGFR-mutated advanced NSCLC. N Engl J Med. 2024;391:1486–98. PubMed
13. Zhao ZR, Lin YB, Ng CS, Zhang R, Wu X, Ou Q, et al. Mutation profile of resected EGFR-mutated lung adenocarcinoma by next-generation sequencing. Oncologist. 2019;24:1368–74. Article PubMed PMC PDF
14. Deng Q, Qiu Y, Jia J, Tang H, Liu L, Huang L, et al. Genetic alteration profile of EGFR-mutant resected IIB-IIIA stage NSCLC and correlation to clinical outcomes. Transl Lung Cancer Res. 2019;8:838–46. Article PubMed PMC
15. Li MM, Datto M, Duncavage EJ, Kulkarni S, Lindeman NI, Roy S, et al. Standards and guidelines for the interpretation and reporting of sequence variants in cancer: a joint consensus recommendation of the Association for Molecular Pathology, American Society of Clinical Oncology, and College of American Pathologists. J Mol Diagn. 2017;19:4–23. PubMed PMC
16. Sanchez-Vega F, Mina M, Armenia J, Chatila WK, Luna A, La KC, et al. Oncogenic signaling pathways in The Cancer Genome Atlas. Cell. 2018;173:321–37. PubMed PMC
17. Lee JS, Kim EK, Kim KA, Shim HS. Clinical impact of genomic and pathway alterations in stage I EGFR-mutant lung adenocarcinoma. Cancer Res Treat. 2024;56:104–14. Article PubMed PMC PDF
18. Zhou J, Sanchez-Vega F, Caso R, Tan KS, Brandt WS, Jones GD, et al. Analysis of tumor genomic pathway alterations using broad-panel next-generation sequencing in surgically resected lung adenocarcinoma. Clin Cancer Res. 2019;25:7475–84. Article PubMed PMC PDF
19. Blakely CM, Watkins TB, Wu W, Gini B, Chabon JJ, McCoach CE, et al. Evolution and clinical impact of co-occurring genetic alterations in advanced-stage EGFR-mutant lung cancers. Nat Genet. 2017;49:1693–704. Article PubMed PMC PDF
20. Guo Y, Song J, Wang Y, Huang L, Sun L, Zhao J, et al. Concurrent genetic alterations and other biomarkers predict treatment efficacy of EGFR-TKIs in EGFR-mutant non-small cell lung cancer: a review. Front Oncol. 2020;10:610923.Article PubMed PMC
21. Ono A, Kenmotsu H, Watanabe M, Serizawa M, Mori K, Imai H, et al. Mutant allele frequency predicts the efficacy of EGFR-TKIs in lung adenocarcinoma harboring the L858R mutation. Ann Oncol. 2014;25:1948–53. Article PubMed
22. Gieszer B, Megyesfalvi Z, Dulai V, Papay J, Kovalszky I, Timar J, et al. EGFR variant allele frequency predicts EGFR-TKI efficacy in lung adenocarcinoma: a multicenter study. Transl Lung Cancer Res. 2021;10:662–74. Article PubMed PMC
23. Leonetti A, Sharma S, Minari R, Perego P, Giovannetti E, Tiseo M. Resistance mechanisms to osimertinib in EGFR-mutated non-small cell lung cancer. Br J Cancer. 2019;121:725–37. Article PubMed PMC PDF
24. Blakely CM, Urisman A, Gubens MA, Mulvey CK, Allen GM, Shiboski SC, et al. Neoadjuvant osimertinib for the treatment of stage I-IIIA epidermal growth factor receptor-mutated non-small cell lung cancer: a phase II multicenter study. J Clin Oncol. 2024;42:3105–14. Article PubMed PMC
25. Li XM, Li WF, Lin JT, Yan HH, Tu HY, Chen HJ, et al. Predictive and prognostic potential of TP53 in patients with advanced non-small-cell lung cancer treated with EGFR-TKI: analysis of a phase III randomized clinical trial (CTONG 0901). Clin Lung Cancer. 2021;22:100–9. Article PubMed
26. Kim Y, Lee B, Shim JH, Lee SH, Park WY, Choi YL, et al. Concurrent genetic alterations predict the progression to target therapy in EGFR-mutated advanced NSCLC. J Thorac Oncol. 2019;14:193–202. Article PubMed
27. Knebel FH, Bettoni F, Shimada AK, Cruz M, Alessi JV, Negrao MV, et al. Sequential liquid biopsies reveal dynamic alterations of EGFR driver mutations and indicate EGFR amplification as a new mechanism of resistance to osimertinib in NSCLC. Lung Cancer. 2017;108:238–41. Article PubMed
28. Jung SJ, Kim DS, Park WJ, Lee H, Choi IJ, Park JY, et al. Mutation of the TERT promoter leads to poor prognosis of patients with non-small cell lung cancer. Oncol Lett. 2017;14:1609–14. Article PubMed PMC
29. Singh RR. Target enrichment approaches for next-generation sequencing applications in oncology. Diagnostics (Basel). 2022;12:1539.Article PubMed PMC
30. Kumbrink J, Demes MC, Jeroch J, Brauninger A, Hartung K, Gerstenmaier U, et al. Development, testing and validation of a targeted NGS-panel for the detection of actionable mutations in lung cancer (NSCLC) using anchored multiplex PCR technology in a multicentric setting. Pathol Oncol Res. 2024;30:1611590.Article PubMed PMC

Figure & Data

REFERENCES

Citations

Citations to this article as recorded by

PubReader
ePub Link

Cite

CITE: export Copy Download Format; Close

Download Citation

Download a citation file in RIS format that can be imported by all major citation management software, including EndNote, ProCite, RefWorks, and Reference Manager.

Format:

RIS — For EndNote, ProCite, RefWorks, and most other reference management software
BibTeX — For JabRef, BibDesk, and other BibTeX-specific software

Include:

Citation for the content below
Citation and abstract for the content below

Genomic Insights into Innate Resistance in Early-Stage EGFR-Mutant Non–Small Cell Lung Cancer: A Comprehensive Analysis of Next-Generation Sequencing Data

Cancer Res Treat. 2026;58(3):738-747. Published online August 25, 2025

DOI: https://doi.org/10.4143/crt.2025.114

XML Download

Abstract
Introduction
Materials and Methods
Results
Discussion
Electronic Supplementary Material
NOTES
REFERENCES

Genomic Insights into Innate Resistance in Early-Stage EGFR-Mutant Non–Small Cell Lung Cancer: A Comprehensive Analysis of Next-Generation Sequencing Data

Fig. 1. Signaling pathway alterations in the overall study population. The figure illustrates the frequency and distribution of genomic alterations classified by canonical oncogenic signaling pathways, including RTK-RAS/MAP kinase, p53, cell cycle, PI3K, Myc, Wnt, Hippo, TGFβ, and Notch. Pathway classification was based on The Cancer Genome Atlas framework. Alterations were identified using targeted next-generation sequencing. MAP, mitogen-activated protein; PI3K, phosphoinositide 3-kinase; RTK, receptor tyrosine kinase; SPA, signaling pathway alteration; TGFβ, transforming growth factor β.

Fig. 2. Tumor mutational burden and variant allele frequency (VAF) of epidermal growth factor receptor (EGFR) mutations in non–small cell lung cancer (NSCLC). The figure compares the tumor mutational burden (TMB) and VAF of EGFR mutations between early- and advanced-stage NSCLC. TMB is expressed as mutations per megabase and VAF represents the proportion of sequencing reads supporting the EGFR mutation.

Fig. 1.

Fig. 2.

Genomic Insights into Innate Resistance in Early-Stage EGFR-Mutant Non–Small Cell Lung Cancer: A Comprehensive Analysis of Next-Generation Sequencing Data

Value	Total (n=160)	Early stage (n=100)	Advanced stage (n=60)	p-value
Age at diagnosis (yr)	68.0 (61-75)	67.0 (59-73)	69.5 (63-75)	0.112
Sex
Female	102 (63.8)	61 (61.0)	41 (68.3)	0.350
Male	58 (36.2)	39 (39.0)	19 (31.7)
Stage
IA	36 (22.5)	36 (36.0)	-	-
IB	25 (15.6)	25 (25.0)	-
IIA	15 (9.4)	15 (15.0)	-
IIB	15 (9.4)	15 (15.0)	-
IIIA	9 (5.6)	9 (9.0)	-
IIIB	7 (4.4)	-	7 (11.7)
IIIC	0	-	0
IV	53 (33.1)	-	53 (88.3)
Specimen collection
Surgical specimen	108 (67.5)	94 (94.0)	14 (23.3)	< 0.001
Small biopsy	52 (32.5)	6 (6.0)	46 (76.7)
Histopathology
Adenocarcinoma	154 (96.3)	97 (97.0)	57 (95.0)	0.197
Adenosquamous cell carcinoma	2 (1.3)	2 (2.0)	0
Squamous cell carcinoma	2 (1.3)	1 (1.0)	1 (1.7)
NSCLC NOS	2 (1.3)	0	2 (3.3)
EGFR mutation
Common EGFR mutation^a)	145 (90.6)	91 (91.0)	54 (90.0)	0.753
EGFR exon 20 insertion	11 (6.9)	6 (6.0)	5 (8.3)
Uncommon EGFR mutation	4 (2.5)	3 (3.0)^b)	1 (1.6)^c)
Smoking status
Never smoker	116 (72.5)	71 (71.0)	45 (75.0)	0.583
Ever smoker	44 (27.5)	29 (29.0)	15 (25.0)
Predominant pattern^d)
Lepidic	5 (4.5)	5 (5.1)	0	0.653
Acinar or papillary	99 (90.0)	89 (89.8)	10 (90.9)
Micropapillary	6 (5.5)	5 (5.1)	1 (9.1)
Differentiation^e)
Well-differentiated	15 (13.5)	15 (15.2)	0	< 0.001
Moderately differentiated	76 (68.5)	71 (71.7)	5 (41.7)
Poorly differentiated	20 (18.0)	13 (13.1)	7 (58.3)

Value	Early stage (n=100)	Advanced stage (n=60)	p-value
No. of patients with concurrent GAs	82 (82.0)	55 (91.7)	0.092
Median No. of concurrent GAs	2 (1-4)	3 (2-9)	0.002
Concurrent SPAs
Low (1-2)	83 (83.0)	36 (60.0)	0.001
High (≥ 3)	17 (17.0)	24 (40.0)
Type of SPA
RTK/RAS/MAP-kinase	100 (100)	60 (100)	N / A
p53	42 (42.0)	33 (55.0)	0.065
Cell cycle	12 (12.0)	19 (31.7)	0.002
PI3K	11 (11.0)	17 (28.3)	0.005
Myc	3 (3.0)	9 (15.0)	0.005
Wnt	3 (3.0)	7 (11.7)	0.041
Hippo	0	3 (5.0)	0.051
TGFβ	0	3 (5.0)	0.051
Notch	1 (1.0)	0	> 0.99
NRF2	0	0	N / A

EGFR mutation subtypes	Early stage (n=100)	Advanced stage (n=60)	p-value
Total subjects	3.8 (2.9-4.8)	3.8 (2.8-5.7)	0.206
Common EGFR mutation	3.8 (2.9-4.8)	4.3 (2.8-5.7)	0.141
EGFR exon 20 insertion	3.3 (2.1-5.9)	2.9 (1.0-2.9)	0.292
Uncommon EGFR mutation	4.8 (3.8-5.7)	7.6 (7.6-7.6)	0.327

EGFR mutation subtypes	Early stage (n=100)	Advanced stage (n=60)	p-value
Total subjects	19.3 (12.5-29.4)	29.6 (18.8-44.8)	0.002
Common EGFR mutation	19.0 (12.5-29.7)	29.9 (19.1-44.9)	< 0.001
EGFR exon 20 insertion	25.7 (16.4-29.4)	14.7 (9.9-21.3)	0.403
Uncommon EGFR mutation	16.8 (13.2-20.6)	32.7 (32.7-32.7)	0.206

Table 1. Baseline characteristics of EGFR-mutant NSCLC

Values are presented as median (interquartile range) or number (%). EGFR, epidermal growth factor receptor; NSCLC NOS, non-small cell lung cancer not otherwise specified.

Concomitant de novo EGFR T790M mutations were found in two patients with early-stage NSCLC and one patient with advanced-stage NSCLC,

One patient with L861Q, one with S768I and G724S, and one with G719A,

One patient with G719A,

Predominant pattern could not be identified in one patient in the early-stage and 49 patients in the advanced-stage,

Differentiation could not be identified in one patient in the early-stage and 48 patients in the advanced-stage.

Table 2. Overall genomic profile of EGFR-mutant NSCLC

Values are presented as number (%) or median (interquartile range). EGFR, epidermal growth factor receptor; GAs, genetic alterations; MAP, mitogen-activated protein; N/A, not applicable; NSCLC, non–small cell lung cancer; PI3K, phosphoinositide 3-kinase; RTK, receptor tyrosine kinase; SPAs, signaling pathway alterations; TGFβ, transforming growth factor β.

Table 3. TMB score

Values of the tumor mutational burden (TMB) score are presented as mutations per megabase, with the median (interquartile range). EGFR, epidermal growth factor receptor.

Table 4. VAF of EGFR mutation

Values of the variant allele frequency (VAF) are presented as percentages, with the median (interquartile range). EGFR, epidermal growth factor receptor.