Circulating Tumor DNA- and Cancer Tissue-Based Next-Generation Sequencing Reveals Comparable Consistency in Targeted Gene Mutations for Advanced or Metastatic Non-Small Cell Lung Cancer
Authors: Weijia Huang1,2, Kai Xu1,2, Zhenkun Liu1,2, Yifeng Wang1,2, Zijia Chen1,2, Yanyun Gao3,4, Renwang Peng3,4, Qinghua Zhou1,2
Affiliations:
- Lung Cancer Center/Lung Cancer Institute, West China Hospital, Sichuan University, Chengdu, Sichuan 610041, China
- Department of Thoracic Surgery, West China Hospital, Sichuan University, Chengdu, Sichuan 610041, China
- Department of General Thoracic Surgery, Inselspital, Bern University Hospital, University of Bern, Bern 3010, Switzerland
- Department for BioMedical Research, University of Bern, Bern 3010, Switzerland
Introduction
Lung cancer remains the leading cause of cancer-related deaths in China, with non-small cell lung cancer (NSCLC) accounting for the majority of cases. While surgical resection improves the prognosis of early-stage NSCLC, targeted therapy and immunotherapy have revolutionized the treatment of advanced or metastatic NSCLC. However, only a quarter of patients benefit from targeted therapy, underscoring the importance of molecular subtyping to guide personalized treatment.
Next-generation sequencing (NGS) is the gold standard for molecular genetic profiling, typically performed on formalin-fixed paraffin-embedded (FFPE) tissue samples. However, tissue biopsies are invasive and may not be feasible for all patients, particularly those with advanced disease or those requiring dynamic monitoring of genetic mutations. Liquid biopsy, specifically the analysis of circulating tumor DNA (ctDNA), has emerged as a promising alternative. Despite its potential, the reliability of ctDNA testing remains controversial, especially in early-stage disease where ctDNA levels are low.
This study aimed to evaluate the consistency of NGS results between ctDNA-based and tissue-based approaches in NSCLC and identify patient characteristics that favor ctDNA testing.
Methods
Patient Enrollment and Sample Acquisition
The study enrolled 85 patients diagnosed with NSCLC at the Lung Cancer Center, West China Hospital, Sichuan University, between December 2017 and August 2022. Both plasma and tumor samples were collected from each patient. Plasma was collected before surgery or systemic treatment, while tumor tissues were obtained from surgical specimens or percutaneous lung puncture biopsies.
Clinical staging was performed according to the eighth edition of the American Joint Committee on Cancer Staging Manual. Clinical data, including age, sex, metastasis, pleural effusion, TNM stage, survival, and serum tumor biomarkers, were extracted from medical records.
Targeted Sequencing and Data Curation
A customized NGS panel targeting 425 cancer-relevant genes was used for sequencing. DNA was extracted from FFPE tissues and plasma samples using QIAamp DNA FFPE Tissue Kit and QIAamp Circulating Nucleic Acid Kit, respectively. Sequencing libraries were constructed using the KAPA HyperDNA Library Prep Kit, and hybridization enrichment was performed using custom probes. The sequencing depth was approximately 5000× for plasma samples and 1000× for tissue samples.
Statistical Analysis
Data were presented as medians and ranges or interquartile ranges (IQR) for quantitative variables and as numbers and percentages for categorical variables. The unweighted Cohen’s kappa coefficient was used to assess the concordance between ctDNA- and tissue-based NGS, with a cutoff value of 0.6 to divide patients into high- and low-concordance groups. Six machine learning models were employed to identify clinical characteristics associated with high concordance.
Results
Patient Characteristics
The study included 85 patients with a median age of 63 years (range: 31–84 years). The majority were male (56.5%) and had lung adenocarcinoma (77.6%). Most patients had advanced disease, with 56.5% diagnosed with stage IV NSCLC. The median tumor mutation burden (TMB) was 4.1 muts/Mb for tissue samples and 5.1 muts/Mb for ctDNA.
Genomic Profiles of ctDNA- and Tissue-Based NGS
Genetic alterations were identified in 82.4% of patients using tissue-based NGS and in 90.6% using ctDNA-based NGS. The most frequently mutated genes in tissue-based NGS were TP53 (49.4%), EGFR (36.5%), STK11 (14.1%), LRP1B (14.1%), and KRAS (11.8%). In ctDNA-based NGS, the most common single-nucleotide polymorphisms (SNPs) were NQO1 (61.2%), GSTM1 (54.1%), XRCC1 (54.1%), and MTHFR (50.6%).
Overall, 51.8% of patients showed consistent gene mutation types between ctDNA- and tissue-based NGS, while 1.2% tested negative in both approaches. Among discordant cases, 8.2% had mutations only in tissue samples, and 17.6% had mutations only in ctDNA.
Association of Clinical Characteristics with Genomic Profiles
In tissue-based NGS, M stage was associated with EGFR and SMARCA4 mutations, while pleural effusion was linked to EGFR mutations. TMB was significantly correlated with mutations in TP53, STK11, KEAP1, and LRP1B. Similar associations were observed in ctDNA-based NGS, with SNPs in NQO1, GSTM1, and XRCC1 showing slight correlations with TNM stage.
Discrimination Ability of Targeted Gene Mutations
The sensitivity, specificity, and accuracy of ctDNA-based NGS for detecting nine targeted genes (TP53, EGFR, KRAS, ALK, MET, ERBB2, BRAF, ROS1, and RET) were high. The kappa test revealed significant differences in stage, bone metastasis, and other organ metastases between high- and low-concordance groups. Patients in the high-concordance group had worse survival outcomes than those in the low-concordance group.
Machine Learning and Prediction Model Visualization
Among six machine learning models, the generalized linear model showed the best prediction accuracy. T stage, M stage, pathological classification, and TMB in ctDNA were the most significant predictors of high concordance. The prediction model achieved an area under the receiver operating characteristic curve (AUC) of 0.79.
Discussion
This study demonstrated that ctDNA-based NGS has comparable detection performance to tissue-based NGS for targeted gene mutations in advanced or metastatic NSCLC. Patients with advanced disease, larger tumor size, and high TMB were more likely to benefit from ctDNA testing.
The high concordance for common driver genes, such as EGFR, ALK, and MET, suggests that ctDNA-based NGS can guide targeted therapy and enable dynamic monitoring of genetic mutations. However, ctDNA-based NGS may not be suitable for early-stage NSCLC due to low ctDNA levels.
The study also highlighted the role of SNPs in genes involved in DNA damage repair and metabolizing enzyme modulation, which may contribute to tumor development and progression. Future research should focus on improving the accuracy and standardization of ctDNA testing to enhance its clinical utility.
Conclusion
ctDNA-based NGS is a reliable alternative to tissue-based NGS for detecting targeted gene mutations in advanced or metastatic NSCLC. Tumor staging and TMB are critical indicators for assessing the credibility of ctDNA-based NGS. This approach offers a less invasive and more dynamic method for guiding personalized cancer treatment.
DOI: doi.org/10.1097/CM9.0000000000003117
Was this helpful?
0 / 0