Skip to main content
NCBI home page
ESMO Open logoLink to ESMO Open
. 2025 Mar 14;10(4):104479. doi: 10.1016/j.esmoop.2025.104479

The impact of co-occurring tumor suppressor mutations with mEGFR as early indicators of relapse in lung cancer

S Hait 1,2, V Noronha 2,3, A Chowdhury 1,2, A Chaudhary 4, B Bawaskar 1, G Dahimbekar 1, S Ahmad 1,2, A Joshi 1,2, V Patil 2,3, N Menon 2,3, M Shah 2,3, R Kaushal 3,5, A Choughule 3, A Bharde 6, J Khandare 6, G Shafi 6, D Lakhwani 1, S Desai 1,2, P Chandrani 2,3,7, K Prabhash 2,3,∗, A Dutt 4,∗
PMCID: PMC11937282  PMID: 40088801

Abstract

Background

Lung adenocarcinoma frequently presents with EGFR mutations, often progressing on EGFR tyrosine kinase inhibitors (TKIs) despite an initial response. Progression is frequently driven by additional genetic changes, including mutations in tumor suppressor genes (TSGs). Understanding the role of these concurrent TSG mutations can help elucidate resistance mechanisms and guide the development of more effective treatment approaches.

Materials and methods

We examined survival outcomes in 483 EGFR-mutant (mEGFR) patients from the GENIE BPC non-small-cell lung cancer (NSCLC) dataset. To understand the mutational landscape and clonal dynamics, whole exome sequencing (WES) was carried out on 48 tumor samples from 16 mEGFR patients at both baseline and post-relapse. A comprehensive gene panel was applied to 200 liquid biopsy samples obtained longitudinally from 25 patients to track clonal evolution.

Results

mEGFR patients with co-occurring TSG mutations exhibited significantly worse outcomes. In the GENIE dataset, overall survival (OS) was shorter [51.11 versus 99.3 months; hazard ratio (HR) 1.8, confidence interval (CI) 1.22-2.75, P = 0.003] and progression-free survival (PFS) was reduced (9.83 versus 11.48 months; HR 1.4, CI 1.03-1.91, P = 0.026). WES analysis revealed 17 TSG mutations that were retained and showed clonal enrichment, particularly in early relapse (progression within 10 months of TKI initiation) or intermediate-stage relapse (relapse occurred between 10 and 20 months), indicated by increased variant allele frequency and their presence was strongly linked to early relapse. Longitudinal clonal studies further confirmed that TSG mutations co-occurring with mEGFR were often truncal, predominantly in early relapsers. Survival analysis using this subset of 17 TSGs showed significantly shorter OS (55.26 versus 99.3 months; HR 1.7, CI 1.12-2.65, P = 0.011) and PFS (9.67 versus 13.12 months; HR 1.5, CI 1.08-2.10, P = 0.013).

Conclusions

A set of 17 co-occurring TSG mutations has been identified as key biomarkers for early relapse in mEGFR lung adenocarcinoma. Longitudinal genomic monitoring, with a focus on clonal evolution, offers valuable insights that can inform personalized treatment strategies and potentially improve patient outcomes.

Key words: lung adenocarcinoma, tyrosine kinase inhibitor, whole exome sequencing, comprehensive gene panel, tumor suppressor genes

Highlights

  • •

    Concurrent TSG mutations and mEGFR correlate with poor OS and PFS.

  • •

    The study identified recurrent TSG mutations linked to relapse that show clonal enrichment with disease progression.

  • •

    Longitudinal liquid biopsy analysis validates co-occurring TSG mutations as truncal, associated with early relapse.

  • •

    The panel of 17 TSG mutations serves as prognostic markers and shows unfavorable outcomes with reduced survival.

  • •

    Longitudinal genome monitoring of clonal evolution can guide personalized therapy to overcome resistance.

Introduction

Lung cancer remains the leading cause of cancer-related mortality worldwide, with non-small-cell lung cancer (NSCLC) accounting for ∼85% of all cases.1,2 Among the various subtypes of NSCLC, lung adenocarcinoma is particularly prevalent, and a significant proportion of Caucasian (∼10%-15%) and Asian (23%-50%) NSCLC patients harbor EGFR-activating mutations, highlighting significant population-based differences in mutation frequency.3, 4, 5, 6 The introduction of EGFR tyrosine kinase inhibitors (TKIs) has revolutionized the treatment landscape for these patients, offering marked improvements in clinical outcomes by specifically targeting these mutations.7, 8, 9 However, despite the initial efficacy, most patients treated with EGFR TKIs eventually experience disease progression or relapse due to the development of acquired resistance. EGFR TKI resistance frequently emerges after a period of clinical response and poses a substantial therapeutic challenge, limiting survival outcomes.10,11 The mechanisms underlying this acquired resistance are complex and multifaceted, often involving concurrent genomic alterations along with the primary EGFR mutations.11,12

Some co-occurring mutations can modulate the tumor’s sensitivity to EGFR TKIs and accelerate the emergence of resistance and disease progression, thereby reducing the overall duration of clinical response.13 A deeper understanding of concurrent genomic alterations is critical as it may unravel predictive biomarkers that are associated with early relapse.14,15 Identifying such biomarkers could guide the development of combinatorial therapeutic strategies designed to counteract preemptive resistance mechanisms and extend EGFR-targeted therapies’ effectiveness.13,16 Additionally, investigating clonal dynamics through longitudinal liquid biopsies offers real-time monitoring of tumor evolution. Circulating tumor DNA (ctDNA) analysis can provide valuable insights into the early molecular events that drive resistance, enabling informed and timely therapeutic interventions.17, 18, 19

Recent studies have highlighted the diverse and heterogeneous mutational landscape of lung adenocarcinomas, particularly in tumors lacking common driver alterations. A study by Carrot-Zhang et al. investigated lung adenocarcinomas without alterations in the RTK/RAS/RAF pathway, identifying significant mutations in several tumor suppressor genes (TSGs), including TP53, STK11, KEAP1, and SMARCA4, as well as focal deletions of TSGs in patients who do not harbor driver mutations such as those in EGFR.20 This in-depth analysis provided new insights into the potential role of TSGs in driving lung cancer progression, especially in the absence of known oncogenic drivers, suggesting that these TSGs could play a crucial role in tumor biology. Similarly, Stockhammer et al. demonstrated that mutational landscapes in patients with early relapse often include co-occurring mutations in TSGs, which may influence the dependency of tumors on EGFR-driven oncogenesis.21 However, the specific impact of these co-occurring TSG mutations in the context of EGFR-mutant (mEGFR) lung adenocarcinoma, particularly in relation to the response to EGFR TKI, remains poorly understood.

Here, we conducted a comprehensive analysis of lung adenocarcinoma patients with mEGFR to explore co-occurring TSG mutations associated with response to EGFR TKI and their role in early relapse. We combined survival data from 483 EGFR-positive patients from the GENIE BPC NSCLC database, whole exome sequencing (WES) of paired tumor samples from 16 patients at both the EGFR TKI-sensitive and -resistant stages, and clonal analysis based on comprehensive genomic profiling of 600 genes from 200 longitudinal liquid biopsy samples from 25 patients. This integrated analysis uncovers how co-occurring mutations in TSGs with EGFR mutations modulate therapeutic response and drive disease relapse to help identify key biomarkers for better management of mEGFR lung adenocarcinoma patients.

Materials and methods

Baseline characteristics of the cohorts

GENIE BPC NSCLC v2.0-public dataset

This dataset includes 483 mEGFR patient samples treated with first-line, first- or second-generation EGFR TKI monotherapy. The patients were treated with erlotinib (n = 99), afatinib (n = 16), or gefitinib (n = 9). To ensure that genetic changes were assessed before any treatment that may affect the outcomes, molecular profiling was carried out at baseline, before first-line therapy was started. To ensure that treatment effects did not skew the survival outcomes, progression-free survival (PFS) and overall survival (OS) studies were limited to patients who underwent baseline molecular testing before starting TKI therapy. The stages of patients included in the BPC cohort were as follows: stage I (4.8%), stage II (17%), stage III (13.2%), stage IV (0.3%), and not available (64.6%).

In-house WES data obtained from 16 NSCLC patients

Patients were categorized by time to relapse as early (<10 months, n = 4), intermediate (10-20 months, n = 7), or late relapse (>20 months, n = 5). First-line treatments were EGFR TKIs for all patients, with gefitinib being the most commonly used agent.

Liquid biopsy cohort

Plasma samples were longitudinally collected from 25 patients with lung adenocarcinoma at baseline and multiple on-treatment time points (total n = 200). Early relapse (<10 months, n = 16), intermediate relapse (10-20 months, n = 6), late relapse (>20 months, n = 1), and no relapse (n = 2) were observed. All patients received gefitinib as the first-line treatment, with 52% receiving combination regimens that included chemotherapy. Among the patients who progressed on EGFR TKIs, 36% continued on gefitinib, 20% were switched to osimertinib, and 32% were treated with chemotherapy alone. The studies involving humans were approved by the Institutional Ethics Committee (IEC) of Tata Memorial Centre (IEC study number: 900233). Written informed consent was obtained from all participants, and the studies were conducted in accordance with the local legislation and institutional requirements.

Survival analysis

OS and PFS analyses were carried out using the GENIE BPC NSCLC v2.0-public dataset, which included 483 mEGFR patient samples.22 The patients were divided into two groups: mEGFR and mEGFR with mTSG. The mEGFR with mTSG group was defined as patients harboring one or more mutations in the top 10 most frequently mutated TSGs identified in the dataset: TP53, CDKN2A, NKX2-1, RB1, RBM10, KMT2D, TERT, APC, ATRX, and SMAD4.

Kaplan–Meier survival curves were generated, and statistical significance was assessed using the log-rank test. To quantify the effect of TSG status on survival outcomes, Cox proportional hazards regression was used to calculate hazard ratios (HRs), providing an estimate of the relative risk of an event (death or progression) in the mEGFR with mTSG group compared with that in the mEGFR group. To strengthen our findings, we carried out a survival analysis on a superset of 17 TSGs created by integrating the top 10 TSGs identified from our WES analysis: RB1, TP53, SMAD4, TET2, BAP1, PTEN, KEAP1, IKZF1, BLM, and NBN, with the 10 most frequently mutated TSGs in the GENIE BPC NSCLC dataset.

For independent validation of our findings, we utilized the MSK-CHORD dataset, which includes 1890 mEGFR patient samples. Similar to the GENIE BPC cohort, patients were classified into two groups: mEGFR and mEGFR with mTSG.23

Tumor samples, sequencing, and bioinformatic analysis

We reanalyzed the existing WES data of formalin-fixed paraffin-embedded EGFR TKI-sensitive and EGFR TKI-resistant tumor samples along with paired blood samples obtained from 16 NSCLC patients treated at the Tata Memorial Centre (unpublished data). FASTQ reads were aligned to the human reference genome (GRCh37) and exon-specific reads were extracted for the EGFR gene locus and converted to a BAM file. After duplicate marking and base recalibration with GATK v4.1.8 (developed by the Broad Institute, Cambridge, MA), two variant callers, HaplotypeCaller and Mutect 2, were employed. Germline variants were excluded by comparison with the matched normal blood samples. Variants were filtered by depth (≥5 reads) using BCFtools (developed by the Wellcome Sanger Institute, Hinxton, UK), and prediction scores were calculated with seven different tools using dbNSFP (Database for Nonsynonymous SNPs' Functional Predictions, developed by the University of Michigan, Ann Arbor, MI). The mutations in the known variants of EGFR were visualized and confirmed using Integrative Genomics Viewer (IGV) v2.8.2 (developed by the Broad Institute, Cambridge, MA). TSGs and oncogenes were classified based on the COSMIC Cancer Gene Census (CGC) (curated by the Wellcome Sanger Institute, Hinxton, UK). Custom approaches were used to identify recurrently altered genes in sensitive and resistant samples, evaluate mutation retention between samples, and calculate changes in variant allele fraction (VAF).

Liquid biopsy collection, sequencing, and bioinformatic analysis

Cell-free DNA (cfDNA) was isolated from patient plasma using a QIAamp MinElute ccfDNA Mini Kit (manufactured by QIAGEN GmbH, Hilden, Germany). Plasma was separated from 10 ml peripheral blood by double centrifugation (2000 g, followed by 3200 g) and cfDNA was extracted as per manufacturer’s protocol. The quantity and quality of cfDNA were assessed using Qubit fluorometry (Invitrogen, a subsidiary of Thermo Fisher Scientific, Waltham, MA) and TapeStation (manufactured by Agilent Technologies, Santa Clara, CA). A total of 200 plasma samples were sequenced using OncoIndex comprehensive gene panel (∼600 genes). Sequencing reads were aligned, and variant calls were generated. A quality check was carried out on these samples to verify the quality of the raw data obtained after sequencing. After trimming, these samples were further subjected to sequencing alignment to the reference genome (hg38) using the BWA MEM tool. For downstream analysis, Sentieon and Mutect2 were used. As these were orphan tumor samples, the tumor-only mode was selected for Mutect2. Variant call files (VCFs) obtained from Mutect2 were further subjected to FilterMutectCalls to select only PASS variants. Furthermore, annotation was done using BCFtools and vep for germline depletion, and the deleterious nature of these variants was determined using functional prediction tool-based analysis of somatic non-synonymous variants using seven different tools. The variants obtained at the end were further considered for validation.

For clonal analysis, the raw VCFs from tumor DNA sequencing were processed into mutation annotation format (MAF) files using the vcf2maf functionality. The MAF files were analyzed to extract key attributes including HUGO symbol, chromosome number, genomic coordinates, reference read counts (ref_count), variant read counts (var_count), and VAFs in tab separated values (TSV) files. Longitudinal sample TSVs from each patient were consolidated for variant calling across time points. Data were formatted as input for PyClone analysis, where the major copy number was taken as 2 and the minor copy number was 0, to deduce clonal population structure and quantify clonal abundance.24 The PyClone output data of cellular prevalence were then used as an input for the ClonEvol tool to analyze and visualize the clonal evolution dynamics of tumors using longitudinal sequencing data.25 It uses a mathematical model to describe clonal dynamics considering processes such as clonal expansion, contraction, and genetic divergence. By analyzing clonal dynamics and fish plot visualizations, ClonEvol provides insights into the clonal composition of the tumor, the emergence of treatment-resistant subclones, and the potential impact of selective pressures on the tumor’s evolutionary landscape.

Results

Co-occurring TSG mutations are associated with decreased survival in EGFR-mutant NSCLC

The survival analysis for both OS and PFS revealed statistically significant differences between the mEGFR with mTSG and mEGFR-only cohorts. Patients harboring co-occurring TSG mutations with EGFR showed considerably worse outcomes compared with those without TSG mutations. The mean OS for mEGFR with mTSG patients was 51.118 months, whereas that for the mEGFR-only patients was 99.30 months. The HR for OS was 1.8, and the confidence interval (CI) was 1.22-2.75 using the mEGFR group as the reference. For PFS, the mean for mEGFR with mTSG patients was 9.83 months, compared with 11.48 months for mEGFR patients, with an HR of 1.4 and a CI of 1.03-1.91, as shown in Figure 1. Kaplan–Meier analysis showed statistically significant differences in both OS (P = 0.003) and PFS (P = 0.026). Building on this finding from the survival outcomes, we suggest that co-occurring TSG mutations may play a crucial role in modulating the response to EGFR TKIs, as patients in the group received TKI treatment. Further analyses, including multivariate approaches, are required to validate this association.

Figure 1.

Figure 1

Open in a new tab

Survival analysis of mEGFR NSCLC patients from the GENIE BPC NSCLC v2.0-public dataset with and without co-occurring TSG mutations. (A) Kaplan–Meier survival curves comparing OS between mEGFR (red) and mEGFR with mTSG (blue) groups. mEGFR with mTSG patients exhibited significantly worse OS (mean OS 51.11 months) compared with mEGFR patients (mean OS 99.30 months). CI 1.22-2.75, P = 0.003. The numbers below the curve indicate patients at risk at each time point. (B) Forest plot showing the HR for OS, with mEGFR as the reference group (HR 1.8). (C) Kaplan–Meier survival curves comparing PFS between mEGFR (red) and mEGFR with mTSG (blue) groups. mEGFR with mTSG patients demonstrated significantly shorter PFS (mean PFS 9.83 months) compared with mEGFR patients (mean PFS 11.48 months). CI 1.03-1.91, P = 0.026. The numbers below the curve indicate patients at risk at each time point. (D) Forest plot showing the HR for PFS, with mEGFR as the reference group (HR 1.4). ∗P value < 0.05, ∗∗P value < 0.01. CI, confidence interval; HR, hazard ratio; mEGFR, EGFR-mutant; NSCLC, non-small-cell lung cancer; OS, overall survival; PFS, progression-free survival; TSG, tumor suppressor gene.

To test this hypothesis further, we carried out WES of primary tumor samples from 16 lung adenocarcinoma patients treated at the Tata Memorial Centre with a mean coverage depth of 100×. Patients were stratified into three categories based on the timing of relapse following initial first-line EGFR TKI therapy: early relapse (<10 months, n = 4), intermediate relapse (10-20 months, n = 7), and late relapse (>20 months, n = 5), as depicted in Figure 2B. Analysis of paired baseline and relapse tumors from these 16 lung adenocarcinoma patients led us to identify the genomic alterations retained during acquired resistance. WES revealed somatic mutations in 1914 unique genes across the cohort. We determined the mutation retention between paired samples. Of the 1914 mutated genes identified, 153 were retained in both the baseline and relapse tumors in one or more patients. Notably, 20 of the recurrently retained genes have been reported to have roles in cancer according to the COSMIC CGC, as shown in Figure 2A. All the patients had activating EGFR mutations (EGFR:pL858R, EGFR exon 19 del) as well as inactivating mutations in TSGs, such as TP53. The role of TP53 mutations as a predictive marker for early relapse on EGFR TKIs has already been reported. The concurrent TP53 mutations in mEGFR NSCLC patients treated with EGFR TKIs significantly impacted the clinical outcomes. Canale et al. showed substantially shorter PFS of 4.2 months compared with 16.8 months in those with wild-type TP53.26 Similarly, this negative prognostic value of TP53 mutations was validated by Vokes et al. This shows that TP53 mutation status can serve as a potential predictor of early relapse in patients receiving EGFR TKI treatment.27 Fourteen of the retained mutated genes with documented roles in cancer were TSGs, whereas only five were oncogenes.

Figure 2.

Figure 2

Open in a new tab

Analysis of co-occurring genomic alterations in mEGFR patients shows enrichment of inactivating mutations in tumor suppressor genes and is associated with early relapse after treatment with TKI. (A) Whole exome sequencing on 48 tumor samples from 16 patients diagnosed with mEGFR lung adenocarcinoma identified mutations in the 1914 genes. Twenty recurrent mutations occurred in known cancer genes (CGC), with significant enrichment for inactivating mutations in tumor suppressor genes (14) compared with activating mutations in oncogenes (5). (B) Non-next generation sequencing detection of baseline EGFR mutations and patients were grouped based on the duration until relapse after treatment: early (<10 months), intermediate (10-20 months), and late (>20 months). (C) Changes in variant allele fractions between paired baseline and resistant specimens are shown for recurrently mutated genes. Increases in VAF in resistant compared with baseline samples were observed for known drivers EGFR and TP53 as well as frequently mutated tumor suppressor genes. These increases were most pronounced in early-relapse and intermediate-relapse subgroups. CGC, Cancer Gene Census; LuAD, lung adenocarcinoma; mEGFR, EGFR-mutant; TKI, tyrosine kinase inhibitor; VAF, variant allele fraction.

We found a significant enrichment of inactivating mutations in key TSG and EGFR mutations, consistent with survival analysis. These included inactivating mutations in TP53, RB1, and BLM, as shown in Figure 2A. Further analysis of the paired baseline and relapse tumors revealed that the early and intermediate relapse groups showed higher retention of TSG mutations and a significant increase in VAF, particularly for mutations in TSGs, as illustrated in Figure 2C. This suggests clonal enrichment under the selective pressure of EGFR inhibitor treatment, where TSG-mutated subclones gain survival advantage and proliferate. Intriguingly, this trend of increasing VAF was not observed in patients with late relapse, possibly due to the greater tumor heterogeneity in these cases, which might dilute the clonal dominance of TSG-mutated subclones. The retention and VAF expansion of co-occurring TSG mutations, particularly in early and intermediate relapsers, strongly suggests that these mutations serve as key drivers of early therapeutic failure. This finding highlights the potential of these TSG mutations as predictive biomarkers for resistance to EGFR TKIs, offering valuable insights for developing strategies to counteract early relapse in mEGFR lung adenocarcinoma patients.

Longitudinal liquid biopsy analysis confirms clonal expansion of the co-occurring TSG mutations following disease relapse

To further validate our findings of survival analysis, primary tumor samples, and track the evolution of resistance mechanisms, we analyzed liquid biopsy samples from 200 plasma samples collected longitudinally from 25 mEGFR patients, as represented in Supplementary Figures S1 and S2, available at https://doi.org/10.1016/j.esmoop.2025.104479. The average sequencing depth was 5000×. Common activating EGFR mutations (L858R, exon 19 deletion) were validated in these 200 longitudinally acquired samples and visualized using IGV, as shown in Supplementary Figure S3, available at https://doi.org/10.1016/j.esmoop.2025.104479. Similar to primary tumors, patients were categorized based on time to relapse after initial first-line EGFR TKI therapy as early relapse (<10 months, n = 16), intermediate relapse (10-20 months, n = 6), late relapse (>20 months, n = 1), or non-relapse (n = 2), as represented in Table 1.

Table 1.

Horizontal sample collection from liquid biopsies of EGFR-mutant lung adenocarcinoma patients

Patient no. EGFR mutation status at baseline No. of horizontal collection Treatment Treatment post-disease progression on TKI Relapse Relapse status
1 L858R 7 Pem/carbo + gef Pem/carbo + gef 8 Early
2 Ex19del 4 Pem/carbo + gef Pem/carbo + gef 9 Early
3 L858R 5 Gefitinib only BSC 9 Early
4 L858R 7 Gefitinib only Pem/carbo 6 Early
5 Ex19del 13 Pem/carbo + gef Pem/carbo + gef 1 Early
6 L858R 5 Gefitinib only Pem/carbo 6 Early
7 L858R 13 Pem/carbo + gef Gemcitabine 9 Early
8 L858R 8 Gefitinib only Pem/carbo 3 Early
9 Ex19del 15 Pem/carbo + gef Pem/carbo + gef 10 Early
10 Ex19del 17 Pem/carbo + gef Osimertinib 2 Early
11 Ex19del 15 Pem/carbo + gef Pacli/carbo 10 Early
12 L858R 8 Pem/carbo + gef Pem/carbo + gef 8 Early
13 Ex19del 5 Gefitinib Gefitinib 7 Early
14 L858R 5 Gefitinib Gefitinib 2 Early
15 Ex19del 5 Gefitinib only Pem/carbo 6 Early
16 L858R 3 Gefitinib only Pem/carbo 6 Early
17 Ex19del 9 Osimertinib + gefitinib Paclitaxel 14 Intermediate
18 Ex19del 4 Gefitinib only Gefitinib only 11 Intermediate
19 Ex19del 3 Pem/carbo + gef Osimertinib 11 Intermediate
20 L858R/T790M 9 Pem/carbo + gef Osimertinib + IT MTX 11 Intermediate
21 L858R 16 Pem/carbo + gef Osimertinib 12 Intermediate
22 Ex19del 6 Gefitinib only Osimertinib 14 Intermediate
23 Ex19del 1 Pem/carbo + gef Pem/carbo/gef 26 Late
24 L858R 14 Gefitinib only Not progressed No No relapse
25 Ex19del 5 Pem/carbo + gef Not progressed No No relapse
Open in a new tab

Plasma samples are acquired from 25 patients subjected to targeted sequencing at multiple time points spanning TKI treatment. EGFR-mutant status by non-next generation sequencing methods. Patients were grouped based on the duration until relapse after treatment: early (<10 months), intermediate (10-20 months), and late (>20 months). Treatment given before and after relapse is tabularized.

Carbo, carboplatin; gef, gefitinib; pacli, paclitaxel; pem, pemetrexed; IT MXT, intrathecal methotrexate; TKI, tyrosine kinase inhibitor.

Targeted next generation sequencing panel sequencing of cfDNA from liquid biopsies yielded a high-depth mutational landscape across longitudinal time points for every patient. Tracking changes in PyClone modeling of VAFs longitudinally revealed the heterogeneity of lung adenocarcinomas and branched subclonal architectures and enabled monitoring of clonal dynamics as the treatment of TKI progressed, as represented in Supplementary Figure S4, available at https://doi.org/10.1016/j.esmoop.2025.104479. In clonal studies of 25 mEGFR lung adenocarcinoma patients, the analysis demonstrated the presence of truncal clones consisting mutations in TSGs. Sixteen of the 25 (64%) patients relapsed early, with disease relapse within 10 months of initiation of EGFR TKI treatment. This high proportion of early relapsers underscores the clinical challenge of acquired resistance in mEGFR NSCLC.21

We curated a set of 17 TSGs, selected from the top 10 genes commonly co-altered with EGFR in both the GENIE BPC NSCLC database and our WES dataset. This set included the genes, TP53, CDKN2A, NKX2-1, RB1, RBM10, KMT2D, TERT, APC, ATRX, SMAD4, TET2, BAP1, PTEN, KEAP1, IKZF1, BLM, and NBN. To ensure consistency and relevance across our analyses, we carried out a clonal evolution analysis using the ClonEvol tool on liquid biopsy samples, focusing primarily on these TSGs. Comparing the early relapsers to intermediate and non-relapsers, as demonstrated in Figure 3 and Supplementary Figure S4, available at https://doi.org/10.1016/j.esmoop.2025.104479, distinct clonal evolution patterns were observed in early-, intermediate-, and no-relapse patients, with the emergence of EGFR T790M mutant clones, diminishing prevalence of TSG-mutated clones, and complex branching patterns indicating rapid tumor evolution observed in early relapsers. These findings align with those of studies showing that early relapsers often harbor more genomic alterations and exhibit a more aggressive tumor biology.21 To further explore the role of TSG mutations in relapse patterns, we analyzed the frequency of mutations in the curated superset of 17 TSGs across early-, intermediate-, and no-relapse cases. Despite the limitations of skewed sample sizes (early: n = 16; intermediate: n = 6; and no relapse: n = 2), a notable pattern emerged. A higher frequency of mutations in the 17 TSGs was observed in early-relapse cases compared with intermediate- and no-relapse cases. This observation supports the hypothesis of increased genomic instability and heterogeneity in early relapsers. Notably, 6 out of 17 TSG mutations in the clones were observed in no-relapse cases. This finding suggests that the mere presence of TSG mutations may not be sufficient to strongly predict relapse, and other factors are likely to contribute to relapse outcomes. Further insights into clonal evolution dynamics associated with different relapse patterns are provided in Supplementary Figure S5, available at https://doi.org/10.1016/j.esmoop.2025.104479, which presents the fish plot and clonal architecture of tumors across the early-, intermediate-, and no-relapse groups, revealing the diverse range of evolutionary trajectories associated with each relapse pattern. In these visualizations, we focused on depicting the most commonly co-altered TSGs to provide a clear and informative representation of the TSG dynamics during treatment, as listed in Table 2. Further studies with larger cohorts are needed to validate these observations and explore additional genomic and molecular factors that influence relapse patterns in NSCLC patients.

Figure 3.

Figure 3

Open in a new tab

Fish plot data of clonal evolution dynamics in lung adenocarcinoma patients with different relapse patterns. (A) Early-relapse patient. The fish plot shows the presence of a set of TSGs maintained in the founder clones across all time points. As resistance emerges, new clones harboring the EGFR T790M mutation appear, and the cellular prevalence of clones with TSGs diminishes over time. (B) Intermediate-relapse patient. The changing trend of cellular prevalence across increasing time points is more static compared with the early-relapse patient. (C) No-relapse patient. The clonal composition and prevalence remain relatively stable over time, with a consistent set of TSGs present in the founder clones. These fish plots highlight the distinct clonal evolution patterns associated with early, intermediate, and no relapse in lung adenocarcinoma patients undergoing targeted therapy. We could not plot fish plots for late-relapse patients as there was only one time point available. (D) Frequency of mutations in a superset of 17 TSGs across early- (n = 16), intermediate- (n = 6), and non-relapse (n = 2) cases. Carbo, carboplatin; osi, osimertinib; MTX, methotrexate; pem, pemetrexed; TSG, tumor suppressor gene.

Table 2.

List of most commonly co-altered TSGs identified in the clonal analysis

Patient no. Relapse Clusters
Founder clone 1 2 3 4 5
1 Early BAP1, CDKN2A, RBM10, TERT, TET2, TP53 BLM
2 Early KMT2D, TET2
3 Early APC, BAP1, KMT2D, NKX2-1, RB1, RBM10, TP53 TET2
4 Early BAP1, CDKN2A, PTEN, APC, KEAP1, KMT2D, RB1, SMAD4, TERT, TET2, TP53
5 Early IKZF1, APC, ATRX, BAP1, CDKN2A, KMT2D, PTEN, RB1, RBM10, SMAD4, TERT, TET2, TP53
6 Early TERT, APC, KEAP1, KMT2D, NBN, NKX2-1, RB1, TET2, TP53
7 Early APC, ATRX, BAP1, BLM, IKZF1, KMT2D, NKX2-1, RB1, TERT, TET2, TP53
8 Early BAP1,APC, ATRX, KMT2D, NBN, PTEN, RB1, RBM10, TERT, TET2, TP53
9 Early ATRX, IKZF1, APC, BAP1, BLM, CDKN2A, KMT2D, NKX2-1, PTEN, RBM10, TERT, TET2 RB1, SMAD4, TP53
10 Early ATRX, BLM, APC, BAP1, IKZF1, KEAP1, KMT2D, PTEN, RB1, RBM10, SMAD4, TERT TET2 TP53
11 Early APC, ATRX, BAP1, CDKN2A, IKZF1, KEAP1, KMT2D, PTEN, RB1, RBM10, SMAD4, TERT, TET2, TP53
12 Early SMAD4, TET2, TP53, APC, IKZF1, KEAP1, KMT2D, PTEN, RB1, TERT
13 Early RBM10, SMAD4, KMT2D, PTEN, RB1, TERT, TET2, TP53
14 Early TERT, KEAP1, PTEN, RBM10 RB1, TP53
15 Early KEAP1, KMT2D, NBN, RB1 TET2
16 Early APC, KMT2D, NKX2-1, RBM10
17 Intermediate TP53 TET2
18 Intermediate RB1, APC, KEAP1, KMT2D, TERT, TET2
19 Intermediate TP53
20 Intermediate SMAD4, TP53,APC, BAP1, KEAP1, KMT2D, RB1, RBM10 TET2 TERT
21 Intermediate ATRX, APC, BAP1, KMT2D, NKX2-1, PTEN, RB1, TERT, TET2, TP53
22 Intermediate BAP1, CDKN2A, IKZF1, KEAP1, KMT2D, PTEN, RB1, TP53 TET2 APC
23 Late RBM10, TET2, TP53
24 No relapse BAP1, TP53, JAK1, SETD2
25 No relapse IKZF1, APC, NKX2-1, RB1, SMAD4, TERT, TET2, TP53
Open in a new tab

The table shows TSGs obtained in the founder clone (cluster 0) and corresponding subclones (clusters 1-4) across the cohort of 25 EGFR-mutant lung adenocarcinoma patients analyzed.

TSG, tumor suppressor gene.

To support our findings, we carried out survival analyses by integrating the top 10 TSGs retained in our WES dataset of 16 paired mEGFR lung adenocarcinoma patients with the 10 most frequently mutated TSGs in the GENIE BPC NSCLC dataset, creating a superset of 17 TSGs, as shown in Figure 4. This analysis corroborated our primary findings of survival analysis, showing significant differences in both OS (HR 1.7, CI 1.12-2.65, P = 0.011) and PFS (HR 1.5, CI 1.08-2.10, P = 0.013) between the mEGFR with mTSG and mEGFR-only groups.

Figure 4.

Figure 4

Open in a new tab

Survival analysis of mEGFR NSCLC patients using a curated panel of 17 TSGs co-occurring with mEGFR in the GENIE BPC NSCLC v2.0-public dataset. (A) Kaplan–Meier survival curves comparing OS between mEGFR (red) and mEGFR with mTSG (blue) groups. mEGFR with mTSG patients exhibited significantly worse OS (mean OS 55.26 months) compared with mEGFR patients (mean OS 99.30 months). CI 1.12-2.65, P = 0.011. The numbers below the curve indicate patients at risk at each time point. (B) Forest plot showing the HR for OS, with mEGFR as the reference group (HR 1.7). (C) Kaplan–Meier survival curves comparing PFS between mEGFR (red) and mEGFR with mTSG (blue) groups. mEGFR with mTSG patients demonstrated significantly shorter PFS (mean PFS 9.67 months) compared with mEGFR patients (mean PFS 13.12 months). CI 1.08-2.10, P = 0.013. The numbers below the curve indicate patients at risk at each time point. (D) Forest plot showing the HR for PFS, with mEGFR as the reference group (HR 1.5). ∗P value < 0.05. CI, confidence interval; HR, hazard ratio; mEGFR, EGFR-mutant; NSCLC, non-small-cell lung cancer; OS, overall survival; PFS, progression-free survival; TSG, tumor suppressor gene.

Additionally, we analyzed the MSK-CHORD dataset, which includes 1890 patients with mEGFR NSCLC. Patients were classified into 2 groups: those with EGFR mutations alone (mEGFR) and those with EGFR mutations accompanied by at least one of the 17 TSG mutations (mEGFR with mTSG). The findings revealed a significant difference in OS between these groups, with the presence of 17 TSG mutations being associated with worse outcomes (HR 2.0, CI 1.7-2.5, P ≤ 0.0001) as shown in Supplementary Figure S6, available at https://doi.org/10.1016/j.esmoop.2025.104479.

To study the impact of co-occurring mutations in TP53 and other TSGs on survival outcomes in patients with mEGFR, we carried out survival analyses using two independent datasets: GENIE BPC NSCLC v2.0-public and MSK-CHORD. Patients were categorized into three groups: (i) mEGFR (mutation in EGFR only), (ii) mEGFR/mTP53 (mEGFR with TP53 mutations only), and (iii) mEGFR/mTSG (mEGFR with mutations in TP53 and additional TSGs). Patients with TP53 mutations alone (mEGFR/mTP53) had significantly worse OS compared with those with EGFR mutations only (HR 1.9, CI 1.55-2.34, P < 0.001). Furthermore, patients with mutations in both TP53 and additional TSGs (mEGFR/mTSG) demonstrated an additive effect, resulting in the worst OS among the groups (HR 2.6, CI 2.16-3.15, P < 0.001). These findings underscore the critical impact of co-occurring TSG mutations in driving poor prognosis in this patient population. PFS analysis was carried out using the GENIE BPC dataset. The trends observed were consistent with those for OS. Patients with TP53 mutations alone (mEGFR/mTP53) had significantly worse PFS compared with the mEGFR-only group (HR 1.5, CI 1.0-2.3, P = 0.03). Notably, the group with mutations in both TP53 and additional TSGs (mEGFR/mTSG) exhibited an additive effect, resulting in even worse PFS (HR 1.9, CI 1.3-2.8, P = 0.0019), as demonstrated in Supplementary Figure S7, available at https://doi.org/10.1016/j.esmoop.2025.104479. These analyses highlight the substantial contribution of 17 TSGs identified in our study to the survival outcomes of mEGFR lung adenocarcinoma patients. The consistent trends across OS and PFS analyses emphasize the critical role of these genetic alterations in determining prognosis and suggest that patients with co-occurring TSG mutations represent a distinct high-risk subgroup.

Collectively, these analyses underscore the importance of truncal TSG mutations as potential drivers of therapeutic resistance, particularly in early relapsers. The distinct evolutionary patterns observed across relapse groups highlight the potential of clonal dynamics as a predictive tool for treatment response and relapse timing.

Discussion

This study outlines an intriguing insight into how co-occurring mutations in TSGs are correlated with the effectiveness of EGFR TKIs and contribute to the early onset of resistance in mEGFR lung adenocarcinoma. The analysis encompasses survival outcomes derived from a large dataset. It integrates detailed genomic profiling of both tumor tissues and serial liquid biopsy samples, underscoring the critical role these TSG mutations play in shaping patient prognosis and therapeutic resistance.

A recent study characterized lung adenocarcinomas lacking RTK/RAS/RAF pathway alterations and highlighted significant mutations in TSGs such as TP53, STK11, KEAP1, and SMARCA4. While their study focused on tumors without common driver alterations, our study demonstrates that these TSG mutations also play an important role in mEGFR tumors, particularly in driving resistance to targeted therapies.20 The impact of co-occurring genomic alterations on patient outcomes in mEGFR lung cancer patients has been the subject of intense studies. Stockhammer et al. reported that co-occurring alterations in multiple TSGs are associated with worse outcomes in mEGFR lung cancer patients.21 Our findings confirm and extend their findings by identifying a broader set of TSGs. The survival analysis of the GENIE BPC NSCLC cohort highlights the clinical impact of co-existing TSG mutations in patients with mEGFR lung adenocarcinoma. Patients harboring both EGFR mutations and concurrent TSG alterations exhibit significantly worse clinical outcomes. This finding suggests that the presence of TSG mutations may serve as a potent negative prognostic indicator, thereby underscoring the need for personalized treatment approaches tailored to this high-risk subgroup.

Using WES of tumor samples from patients with mEGFR, we identified loss-of-function mutations in key TSGs, including TP53, RB1, and PTEN, as critical contributors to resistance against TKI therapy and early disease relapse. We thus describe a curated set of 17 TSGs frequently co-altered with mEGFR to emphasize our focus on the most impactful TSGs. This panel combines the top 10 most frequently mutated tumor suppressors from the GENIE BPC NSCLC dataset (TP53, CDKN2A, NKX2-1, RB1, RBM10, KMT2D, TERT, APC, ATRX, and SMAD4) with additional TSGs identified through our WES analysis (TET2, BAP1, PTEN, KEAP1, IKZF1, BLM, and NBN). The findings of survival analysis across the GENIE BPC datasets and our WES gene sets strongly point toward the association of co-occurring TSG mutations in mEGFR NSCLC patients with worse clinical outcomes. This highlights the importance of comprehensive genomic profiling in NSCLC patients, as the identification of co-occurring TSG mutations could have significant implications for prognosis and potentially guide treatment decisions. The expansion in VAF observed in patients with early relapse indicates the selective outgrowth of TSG-mutated subclones, which likely drives the acquisition of resistance.

Our longitudinal analysis of liquid biopsy samples from mEGFR lung adenocarcinoma patients has yielded significant insights into the clonal evolution and emergence of resistant clones. The role of ctDNA analysis for tracking early lung cancer metastatic dissemination and its utility in minimal residual disease detection are well established.19,28 Kobayashi and Tan used single-cell sequencing to unmask intratumor heterogeneity and clonal evolution in mEGFR NSCLC.29 While they focused on spatial heterogeneity, our study adds insights into temporal evolution through liquid biopsy analysis. The findings from liquid biopsy analysis not only validate our primary findings in survival analysis but also provide a deeper understanding of the complex dynamics underlying treatment response and relapse. By leveraging targeted sequencing of ctDNA in plasma samples from longitudinal liquid biopsies, we traced the evolutionary trajectories of subclones contributing to tumor relapse. The high proportion of early relapsers (64%) in our cohort highlights the clinical challenge of acquired resistance in mEGFR NSCLC at an early stage.21 Clonal analysis, based on the VAF of mutations, revealed how the subclonal architecture of tumors evolves under the selective pressure exerted by sequential therapies, as described earlier.30 PyClone modeling provided further insights, revealing recurrent truncal mutations in TSGs across 25 patients, highlighting the robustness of these alterations in tumor biology. Our clonal evolution analysis in patient 6 reveals the emergence of EGFR T790M mutations post-resistance, alongside the presence of TSG mutations in the founding clone. Consistency across our analyses was maintained by focusing on a curated set of 17 TSGs frequently co-altered with EGFR, drawn from both the AACR GENIE BPC database and our WES dataset. This standardized approach, visualized through fish plots, facilitated the interpretation of clonal evolution across patient samples. The clonal evolution patterns we observed through longitudinal liquid biopsy analysis contribute to our understanding of the dynamics of lung cancer resistance, though further validation in larger cohorts is necessary. This approach corroborates recent studies emphasizing the importance of longitudinal cfDNA analysis in capturing tumor evolution and resistance mechanisms.31,32 Chabon et al. used a similar approach by integrating genomic features for early detection.33 Our study specifically focuses on the evolution of resistance in mEGFR tumors under the selective pressure of TKI therapy, providing a more targeted perspective. As represented in Figure 3, distinct clonal evolution patterns were observed in early, intermediate, and late relapsers. Early relapsers exhibited more complex branching patterns and rapid tumor evolution with frequent mutation in TSGs, aligning with studies showing that these patients often harbor more genomic alterations and exhibit more aggressive tumor biology.34 These patterns could potentially serve as prognostic indicators, allowing clinicians to predict the likelihood and timing of relapse and enabling more proactive treatment adjustments.

The persistence of truncal TSG mutations in progenitor clones emphasizes their potential role in tumor initiation and relapse, likely through Darwinian selection mechanisms.35,36 Tumors harboring pre-existing mutations in TSG—or those that acquire such mutations early during treatment—may gain a fitness advantage that enables them to withstand therapeutic pressures and proliferate. This evolutionary insight underscores the importance of assessing tumor heterogeneity and recognizing the presence of resistant subclones when developing therapeutic strategies.13 By demonstrating the feasibility of tracking clonal dynamics through ctDNA analysis, we provide further support for integrating liquid biopsies into clinical practice for better personalized treatment approaches.17,37 Studies have shown that the expression of large tumor suppressor kinases regulates chemotherapy response in NSCLC.38 This underscores that the role of TSGs in treatment outcomes is applicable across different therapeutic modalities. For patients identified with multiple TSG mutations, especially in the truncal clone, clinicians might consider more aggressive or combination treatment approaches from the outset, rather than relying solely on EGFR TKIs. This tailored approach could potentially delay or prevent early relapse in high-risk patients.39,40 The findings illustrate the power of longitudinal sampling in capturing the heterogeneity and clonal dynamics of mEGFR lung adenocarcinoma. Early detection of TSG mutations may help identify patients who are likely to relapse early on TKIs and can go for alternative treatment sequencing or combination therapies. Furthermore, the power of longitudinal sampling in capturing the heterogeneity and clonal dynamics of mEGFR lung adenocarcinoma emphasizes the potential of liquid biopsy as a non-invasive tool for real-time monitoring of treatment response and resistance emergence.

Our study provides novel insights into resistance mechanisms and potential avenues for personalized treatment strategies, but we acknowledge certain limitations that may influence the interpretation of our results. Firstly, the statistical power to assess the impact of individual TSGs is limited, potentially reducing the resolution of our analyses. Future studies with larger sample sizes and more comprehensive datasets may address this limitation and further validate our findings. Secondly, the lack of multivariate analyses in some aspects of the study could have limited our ability to account for potential confounding variables. Incorporating multivariate approaches in subsequent investigations would enhance the robustness of the conclusions and better elucidate complex interactions between genetic and clinical factors. Thirdly, the small subsets of samples analyzed via WES may have constrained our ability to detect rare genetic events. However, the consistency of findings across these subsets and the concordance with previous literature underscore the reliability of our observations. Expanding the cohort size for WES analyses in future work will allow for more detailed exploration of rare mutational landscapes. Finally, the heterogeneous use of first-line treatment modalities among the study cohort could have introduced potential variability in treatment outcomes. While this reflects real-world clinical diversity, standardizing treatment regimens or stratifying analyses based on specific modalities in future studies may yield additional insights into the interplay between treatment and molecular profiles. Despite these limitations, our study represents a significant contribution by integrating longitudinal genomic monitoring with clonal evolution analysis in identifying a distinct set of 17 co-occurring TSG mutations as key biomarkers of early relapse in mEGFR lung adenocarcinoma.

Additionally, while the 17 TSG variants were cross-referenced with tumor tissue sequencing data to confirm their presence and retention during disease progression—establishing them as somatic and demonstrating dynamic co-evolution with clinical disease—there remains a possibility of misclassifying TP53 and TET2 mutations, which overlap with known clonal hematopoiesis of indeterminate potential-associated genes, in few patients, as somatic variants.41 This misclassification could potentially confound the interpretation of liquid biopsy results.

Overall, our integrated approach, combining survival data with tissue and liquid biopsy genomic analysis, offers significant translational potential.42 The identification of 17 TSG mutations as biomarkers to predict early relapse in mEGFR lung adenocarcinoma has significant clinical implications. Early detection of these mutations could enable a more personalized approach to treatment, allowing clinicians to identify high-risk patients and adjust therapeutic strategies accordingly, potentially incorporating additional or alternative therapies to counteract emerging resistance.33,43 The ability to detect emerging resistant clones before clinical relapse creates opportunities for early intervention.44 Profiling TSG mutations can enhance prognostic modeling, guiding the sequencing and combination of therapies for patients with EGFR-positive lung adenocarcinoma. The potential of combining EGFR inhibitors with therapies targeting TSG-driven resistance mechanisms merits further exploration.45 Additionally, longitudinal monitoring through liquid biopsies offers a non-invasive method to track these clonal dynamics in real time, ultimately improving patient outcomes by guiding more effective, individualized treatment strategies.17

Acknowledgements

The authors would like to thank all the technical staff of the Department of Medical Oncology and Department of Pathology, Tata Memorial Hospital/ACTREC, Mumbai, India, who assisted in this study. This research was supported by the Science and Engineering Research Board (SERB), Government of India, under grant EMR/2016/007218, and Indian Council of Medical Research (ICMR), Government of India, under grant 01/ICMRCAREP-2023-0000278/AD/DevDiv. We gratefully acknowledge SERB and ICMR for their financial support, which was instrumental in conducting this study.

Funding

This work was supported by a collaborative basic research project involving the Tata Memorial Centre, an autonomous grant-in-aid institution under the aegis of the Department of Atomic Energy (DAE), Government of India, the Science and Engineering Research Board (SERB), Government of India [grant number EMR/2016/007218], the Indian Council of Medical Research [grant number 01/ICMRCAREP-2023-0000278/AD/DevDiv], and OneCell Diagnostics Pvt Ltd, a commercial entity. This collaboration was established through an institutional Memorandum of Understanding (MOU). OneCell Diagnostics Pvt Ltd had the following involvement with the study: DNA/RNA extraction from liquid biopsy, customized capturing of target regions was generated using proprietary 600 odd gene panel, and sequencing on the Illumina platform to generate sequencing data, which was subsequently independently analyzed by the authors of the study at the Tata Memorial Centre. No financial transactions took place between OneCell Diagnostics Pvt Ltd and the Tata Memorial Centre. None of the funders had any role in the analysis and interpretation of data in the study.

Disclosure

AB, JK, and GS are employed by OneCell Diagnostics Pvt Ltd. All other authors have declared no conflicts of interest.

Contributor Information

K. Prabhash, Email: kumarprabhashtmh@gmail.com.

A. Dutt, Email: amitdutt@south.du.ac.in.

Supplementary data

Supplementary Figure 1
mmc1.pdf (1.3MB, pdf)
Supplementary Figure 2
mmc2.pdf (481.2KB, pdf)

References

  • 1.Sung H., Ferlay J., Siegel R.L., et al. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2021;71(3):209–249. doi: 10.3322/caac.21660. [DOI] [PubMed] [Google Scholar]
  • 2.Noronha V., Budukh A., Chaturvedi P., et al. Uniqueness of lung cancer in Southeast Asia. Lancet Reg Health Southeast Asia. 2024;27 doi: 10.1016/j.lansea.2024.100430. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Skoulidis F., Heymach J.V. Co-occurring genomic alterations in non-small-cell lung cancer biology and therapy. Nat Rev Cancer. 2019;19(9):495–509. doi: 10.1038/s41568-019-0179-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Chandrani P., Prabhash K., Prasad R., et al. Drug-sensitive FGFR3 mutations in lung adenocarcinoma. Ann Oncol. 2017;28(3):597–603. doi: 10.1093/annonc/mdw636. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Chougule A., Prabhash K., Noronha V., et al. Frequency of EGFR mutations in 907 lung adenocarcioma patients of Indian ethnicity. PLoS One. 2013;8(10) doi: 10.1371/journal.pone.0076164. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Soo R.A., Reungwetwattana T., Perroud H.A., et al. Prevalence of EGFR mutations in patients with resected stages I to III NSCLC: results from the EARLY-EGFR study. J Thorac Oncol. 2024;19(10):1449–1459. doi: 10.1016/j.jtho.2024.06.008. [DOI] [PubMed] [Google Scholar]
  • 7.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(9):725–737. doi: 10.1038/s41416-019-0573-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Choughule A., Sharma R., Trivedi V., et al. Coexistence of KRAS mutation with mutant but not wild-type EGFR predicts response to tyrosine-kinase inhibitors in human lung cancer. Br J Cancer. 2014;111(11):2203–2204. doi: 10.1038/bjc.2014.401. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Noronha V., Prabhash K., Thavamani A., et al. EGFR mutations in Indian lung cancer patients: clinical correlation and outcome to EGFR targeted therapy. PLoS One. 2013;8(4) doi: 10.1371/journal.pone.0061561. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Mok T.S., Wu Y.L., Thongprasert S., et al. Gefitinib or carboplatin-paclitaxel in pulmonary adenocarcinoma. N Engl J Med. 2009;361(10):947–957. doi: 10.1056/NEJMoa0810699. [DOI] [PubMed] [Google Scholar]
  • 11.Chmielecki J., Mok T., Wu Y.L., et al. Analysis of acquired resistance mechanisms to osimertinib in patients with EGFR-mutated advanced non-small cell lung cancer from the AURA3 trial. Nat Commun. 2023;14(1):1071. doi: 10.1038/s41467-023-35962-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Oxnard G.R., Hu Y., Mileham K.F., et al. Assessment of resistance mechanisms and clinical implications in patients with EGFR T790M-positive lung cancer and acquired resistance to osimertinib. JAMA Oncol. 2018;4(11):1527–1534. doi: 10.1001/jamaoncol.2018.2969. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Blakely C.M., Watkins T.B.K., Wu W., et al. Evolution and clinical impact of co-occurring genetic alterations in advanced-stage EGFR-mutant lung cancers. Nat Genet. 2017;49(12):1693–1704. doi: 10.1038/ng.3990. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Hartmaier R.J., Albacker L.A., Chmielecki J., et al. High-throughput genomic profiling of adult solid tumors reveals novel insights into cancer pathogenesis. Cancer Res. 2017;77(9):2464–2475. doi: 10.1158/0008-5472.CAN-16-2479. [DOI] [PubMed] [Google Scholar]
  • 15.Chougule A., Chandrani P., Noronha V., et al. Real-world evidence of EGFR targeted therapy in NSCLC - a brief report of decade long single center experience. JTO Clin Res Rep. 2023;4(11) doi: 10.1016/j.jtocrr.2023.100566. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Besse B., Pons-Tostivint E., Park K., et al. Biomarker-directed targeted therapy plus durvalumab in advanced non-small-cell lung cancer: a phase 2 umbrella trial. Nat Med. 2024;30(3):716–729. doi: 10.1038/s41591-024-02808-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Siravegna G., Mussolin B., Venesio T., et al. How liquid biopsies can change clinical practice in oncology. Ann Oncol. 2019;30(10):1580–1590. doi: 10.1093/annonc/mdz227. [DOI] [PubMed] [Google Scholar]
  • 18.Abbosh C., Birkbak N.J., Swanton C. Early stage NSCLC - challenges to implementing ctDNA-based screening and MRD detection. Nat Rev Clin Oncol. 2018;15(9):577–586. doi: 10.1038/s41571-018-0058-3. [DOI] [PubMed] [Google Scholar]
  • 19.Abbosh C., Frankell A.M., Harrison T., et al. Tracking early lung cancer metastatic dissemination in TRACERx using ctDNA. Nature. 2023;616(7957):553–562. doi: 10.1038/s41586-023-05776-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Carrot-Zhang J., Yao X., Devarakonda S., et al. Whole-genome characterization of lung adenocarcinomas lacking alterations in the RTK/RAS/RAF pathway. Cell Rep. 2021;34(8) doi: 10.1016/j.celrep.2021.108784. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Stockhammer P., Grant M., Wurtz A., et al. Co-occurring alterations in multiple tumor suppressor genes are associated with worse outcomes in patients with EGFR-mutant lung cancer. J Thorac Oncol. 2024;19(2):240–251. doi: 10.1016/j.jtho.2023.10.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Choudhury N.J., Lavery J.A., Brown S., et al. The GENIE BPC NSCLC cohort: a real-world repository integrating standardized clinical and genomic data for 1,846 patients with non-small cell lung cancer. Clin Cancer Res. 2023;29(17):3418–3428. doi: 10.1158/1078-0432.CCR-23-0580. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Jee J., Fong C., Pichotta K., et al. Automated real-world data integration improves cancer outcome prediction. Nature. 2024;636(8043):728–736. doi: 10.1038/s41586-024-08167-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Roth A., Khattra J., Yap D., et al. PyClone: statistical inference of clonal population structure in cancer. Nat Methods. 2014;11(4):396–398. doi: 10.1038/nmeth.2883. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Dang H.X., White B.S., Foltz S.M., et al. ClonEvol: clonal ordering and visualization in cancer sequencing. Ann Oncol. 2017;28(12):3076–3082. doi: 10.1093/annonc/mdx517. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Canale M., Petracci E., Delmonte A., et al. Impact of TP53 mutations on outcome in EGFR-mutated patients treated with first-line tyrosine kinase inhibitors. Clin Cancer Res. 2017;23(9):2195–2202. doi: 10.1158/1078-0432.CCR-16-0966. [DOI] [PubMed] [Google Scholar]
  • 27.Vokes N.I., Chambers E., Nguyen T., et al. Concurrent TP53 mutations facilitate resistance evolution in EGFR-mutant lung adenocarcinoma. J Thorac Oncol. 2022;17(6):779–792. doi: 10.1016/j.jtho.2022.02.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Assaf Z.J.F., Zou W., Fine A.D., et al. A longitudinal circulating tumor DNA-based model associated with survival in metastatic non-small-cell lung cancer. Nat Med. 2023;29(4):859–868. doi: 10.1038/s41591-023-02226-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Kobayashi K., Tan A.C. Unraveling the impact of intratumoral heterogeneity on EGFR tyrosine kinase inhibitor resistance in EGFR-mutated NSCLC. Int J Mol Sci. 2023;24(4):4126. doi: 10.3390/ijms24044126. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Dagogo-Jack I., Shaw A.T. Tumour heterogeneity and resistance to cancer therapies. Nat Rev Clin Oncol. 2018;15(2):81–94. doi: 10.1038/nrclinonc.2017.166. [DOI] [PubMed] [Google Scholar]
  • 31.Pasquale R., Forgione L., Roma C., et al. Targeted sequencing analysis of cell-free DNA from metastatic non-small-cell lung cancer patients: clinical and biological implications. Transl Lung Cancer Res. 2020;9(1):61–70. doi: 10.21037/tlcr.2020.01.01. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Jung H.A., Ku B.M., Kim Y.J., et al. Longitudinal monitoring of circulating tumor DNA from plasma in patients with curative resected stages I to IIIA EGFR-mutant non-small cell lung cancer. J Thorac Oncol. 2023;18(9):1199–1208. doi: 10.1016/j.jtho.2023.05.027. [DOI] [PubMed] [Google Scholar]
  • 33.Chabon J.J., Hamilton E.G., Kurtz D.M., et al. Integrating genomic features for non-invasive early lung cancer detection. Nature. 2020;580(7802):245–251. doi: 10.1038/s41586-020-2140-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Dentro S.C., Leshchiner I., Haase K., et al. Characterizing genetic intra-tumor heterogeneity across 2,658 human cancer genomes. Cell. 2021;184(8):2239–2254.e39. doi: 10.1016/j.cell.2021.03.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Turajlic S., Sottoriva A., Graham T., Swanton C. Resolving genetic heterogeneity in cancer. Nat Rev Genet. 2019;20(7):404–416. doi: 10.1038/s41576-019-0114-6. [DOI] [PubMed] [Google Scholar]
  • 36.McGranahan N., Swanton C. Clonal heterogeneity and tumor evolution: past, present, and the future. Cell. 2017;168(4):613–628. doi: 10.1016/j.cell.2017.01.018. [DOI] [PubMed] [Google Scholar]
  • 37.Wang W., He Y., Yang F., Chen K. Current and emerging applications of liquid biopsy in pan-cancer. Transl Oncol. 2023;34 doi: 10.1016/j.tranon.2023.101720. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Luo S.Y., Kwok H.H., Yang P.C., Ip M.S., Minna J.D., Lam D.C. Expression of large tumour suppressor (LATS) kinases modulates chemotherapy response in advanced non-small cell lung cancer. Transl Lung Cancer Res. 2020;9(2):294–305. doi: 10.21037/tlcr.2020.03.26. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Soria J.C., Ohe Y., Vansteenkiste J., et al. Osimertinib in untreated EGFR-mutated advanced non-small-cell lung cancer. N Engl J Med. 2018;378(2):113–125. doi: 10.1056/NEJMoa1713137. [DOI] [PubMed] [Google Scholar]
  • 40.Ortiz-Cuaran S., Scheffler M., Plenker D., et al. Heterogeneous mechanisms of primary and acquired resistance to third-generation EGFR inhibitors. Clin Cancer Res. 2016;22(19):4837–4847. doi: 10.1158/1078-0432.CCR-15-1915. [DOI] [PubMed] [Google Scholar]
  • 41.Chan H.T., Chin Y.M., Nakamura Y., Low S.K. Clonal hematopoiesis in liquid biopsy: from biological noise to valuable clinical implications. Cancers (Basel) 2020;12(8):2277. doi: 10.3390/cancers12082277. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Ito S., Tsurumi K., Shindo N., et al. Multi-gene liquid biopsy to detect resistance to first-line osimertinib in patients with EGFR-mutated lung adenocarcinoma. Anticancer Res. 2023;43(11):5031–5040. doi: 10.21873/anticanres.16702. [DOI] [PubMed] [Google Scholar]
  • 43.Guo H., Zhang J., Qin C., et al. Biomarker-targeted therapies in non-small cell lung cancer: current status and perspectives. Cells. 2022;11(20):3200. doi: 10.3390/cells11203200. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Gale D., Heider K., Ruiz-Valdepenas A., et al. Residual ctDNA after treatment predicts early relapse in patients with early-stage non-small cell lung cancer. Ann Oncol. 2022;33(5):500–510. doi: 10.1016/j.annonc.2022.02.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Shi K., Wang G., Pei J., et al. Emerging strategies to overcome resistance to third-generation EGFR inhibitors. J Hematol Oncol. 2022;15(1):94. doi: 10.1186/s13045-022-01311-6. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Figure 1
mmc1.pdf (1.3MB, pdf)
Supplementary Figure 2
mmc2.pdf (481.2KB, pdf)

RESOURCES