Genomic profiling of NGS-based ctDNA from Chinese non-small cell lung cancer patients

Cell-free circulating tumor DNA (ctDNA) in plasma enables rapid and repeat testing of actionable mutations. Next-generation sequencing (NGS) is an attractive platform for multiplex sequencing capabilities compared to traditional methods such as PCR. The purpose of this study is to evaluate the value of the NGS-based ctDNA assay and to identify the genomic alteration profile of ctDNA in real-world Chinese non-small cell lung (NSCLC) patients. In total, 294 Chinese patients with pathological diagnosis of Phase III-IV NSCLC were enrolled. 3–4 mL peripheral blood was collected and NGS-based analysis was carried out using a 20-gene panel. The analytical sensitivity and specificity of ctDNA NGS-based assay was validated using droplet digital PCR (ddPCR). We have tested 570 sites from 286 samples using ddPCR, which included 108 positive sites and 462 negative sites from NGS results, and the concordance rate was 99.8% (418/419) for single-nucleotide variants (SNVs) and 96.7% (146/151) for insertions and deletions (InDels). The most frequent genes were TP53 (32%), EGFR (31.97%), KRAS (6.46%), PIK3CA (4.76%), and MET (4.08%). Exon 19 deletion (19del) was the most common alteration in EGFR and G12C was the most common alteration in KRAS. Furthermore, the detection rate of TP53 was higher in the male and patients with squamous cell carcinoma. We also found the prevalence of TP53 in L858R was higher than in 19del (61.29% vs. 40%; p = 0.1115). The results indicate that the results of NGS-based ctDNA assay are highly consistent with ddPCR. In Chinese NSCLC patients, TP53 mutation was more frequently associated with male and squamous cell carcinoma. The prevalence of concomitant mutations in L858R may be different from that in 19del.


Introduction
Non-small cell lung cancer (NSCLC), the most prevalent type of lung cancer, is one of the leading causes of cancerrelated mortality worldwide (Gobbini et al. 2020). Targeted Yanfeng Xi, Zhongyuan Bai and Sihang Gao have contributed equally to this work. therapy has shown great effectiveness for NSCLC patients nowadays (Chen et al. 2020). The National Comprehensive Cancer Network (NCCN) guidelines recommend broad molecular profiling, including specific alterations in EGFR, ALK, ROS1, MET, ERBB2, and RET. Based on these specific alterations, several targeted agents can be used in the selection of optimal therapy strategies (Meng et al. 2019). In addition, various NSCLC patients develop the disease from different risk factors, both genetic and environmental. Therefore, each patient has its unique driver mutation profile . Molecular profiling of the individual tumor genomes can shed light on the distinct molecular mechanisms that regulate cancer progression and facilitate the discovery of potential therapeutic targets.
Traditionally, tissue-based genomic analysis has been the gold standard of precision medicine. However, tumor tissue sampling is often unfeasible for various reasons (Rolfo et al. 2018). Approximately 30% of the patients do not have access to tissue samples for biomarker testing because invasive biopsy procedures may present a health risk (Esposito Abate et al. 2020). However, tumor DNA can be released into the bloodstream from circulating tumor cells (CTCs), primary tumors, minimal residual disease (MRD) and metastatic sites (Li and Liang 2020). Several studies have illustrated that it is practicable to assess tumor mutation status by using circulating free tumor-derived DNA (ctDNA), which is undoubtedly the regular clinical practice of a liquid biopsy in patients with lung cancer (Cho et al. 2022;Pecuchet et al. 2016;Remon et al. 2019). Moreover, ctDNA testing can monitor the genetic status of the cancer in real-time during treatment to identify the mechanisms of resistance. In addition, ctDNA testing can provide a more comprehensive picture of tumor heterogeneity than tissue biopsy ) (Scilla and Rolfo 2019).
PCR-based methods, such as Droplet digital PCR (ddPCR), are the most commonly used technologies to detect the variants in ctDNA (Jiang et al. 2020). However, those technologies are usually limited to the detection of specific defined mutations in a few genes, ignoring the full range of mutations that can occur in the context of acquired drug resistance during targeted therapy (Feng et al. 2018). Amplicon-based next-generation sequencing (NGS) allows the detection of a wider range of variants, which makes it possible to identify novel resistance mutations (Ding et al. 2019). Besides, NGS has demonstrated excellent performance, great compatibility with low input DNA and low cost. These advantages make NGS a promising method for NSCLC detection (Zhang et al. 2017).
In this study, we aim to validate the performance of NGS-based analysis and characterize the genomic alteration profile of ctDNA in patients with NSCLC in China using a 20-gene NGS panel. This study emphasizes the importance of plasma-based NGS ctDNA testing in all NSCLC patients to identify actionable mutations amenable to targeted therapy.

Patient enrollment
This cohort included 294 Chinese NSCLC patients from Shanxi Cancer Hospital (Shanxi, China) from December 2020 to April 2021. This study was approved by the ethics committee of the Shanxi Cancer Hospital in accordance with the provisions of the Declaration of Helsinki. Written informed consent was obtained from each patient for the use of their plasma samples. Clinical characteristics were collected from medical records, including age, gender, clinical stage, and pathology. Patients were selected based on the following criteria: (1) age of 18 or above; (2) diagnosed as stage III or IV lung cancer; (3) a peripheral blood sample of 14-20 mL can be collected; (4) voluntary informed consent. Patients were excluded according to the following criteria: (1) with unqualified nucleic acid quality after DNA extraction; (2) diagnosed with a second primary malignant tumor; (3) who had previously received transplant surgery; (4) who had received an allogeneic blood transfusion or immunotherapy that may introduce foreign DNA.

Sample processing and plasma DNA extraction
Peripheral blood was collected and separated into plasma and buffy coat by centrifugation at 4 ℃ for 10 min at 1600g, followed by centrifugation at 16,000g for 10 min. cfDNA extraction from 3-4 mL of plasma was then performed using the Nucleic Acid Extraction Kit (Beijing USCI Medical Devices Co., Ltd., TQ003, China), following the manufacturer's instructions. The concentration of DNA was determined by Qubit dsDNA high-sensitivity (HS) assay kit (Invitrogen, Q32854, USA).

Capture-based targeted DNA sequencing
Libraries were generated from over 20 ng of cfDNA using Human EGFR, KRAS, BRAF, NRAS, PIK3CA Gene Mutation Joint Detection Kit (Combinatorial Probe Anchor Synthesis) (Beijing USCI Medical Devices Co., Ltd., JK001, China), following the manufacturer's instructions. The libraries were then hybridized to a tailored 20-gene panel including EGFR, ALK, ROS1, etc. based on the gene mutation distribution and gene target drug sites in NSCLC. This panel covers a total of 92 kb and enables one to analyze > 150 SNVs and short InDels that are frequently mutated in NSCLC. PCR products were measured using an Agilent 2100 bioanalyzer and ABI 7500 real-time PCR system (Life Technologies, 4351107, USA) to estimate insert size and concentration. After quality control, the pooled library was sequenced on the USCISEQ-200 analyzers (Beijing USCI Medical Devices Co., Ltd., China) to generate pair-end reads of 100 bp.

Droplet digital PCR
The primers and TaqMan hydrolysis probes for ddPCR were designed by Bio-Rad (USA). For each INDEL and SNP, one probe was labeled with FAM and the opposite probe was labeled with HEX. ddPCR reaction mixture was prepared as follows: 10 μL ddPCR Supermix for Probes (Bio-Rad), 1 μL Primers & Probe (FAM labeled + HEX labeled + Primers) and about 20 ng fragmented nucleic acid. Then the reaction system was supplemented with Nuclease free water to 20 μL. The ddPCR amplification of target region was performed using C1000 Thermal Cycler (Bio-Rad). After amplification, samples were loaded into the OX200 Droplet Reader (Bio-Rad). The final evaluation was performed by QuantaSoft software (version 1.7.4.0917, Bio-Rad).

Patient characteristics
In this study, 294 NSCLC patients' plasma samples were collected, all of whom had a successful NGS ctDNA test performed. The demographic and clinical characteristics of the patients were summarized in Table 1

Prevalence of EGFR and TP53 with different characteristics
We then explored the detection rate of EGFR and TP53 with different characteristics, including gender and pathological types ( Fig. 3a and b). Across different subgroups, the detection rate of EGFR was higher in women and patients with adenocarcinoma. In turn, subgroups of patients who were male and squamous cell carcinoma had a higher rate of mutations in the TP53 gene. Notably, there is a statistically significant difference in the ratio of males and females in the squamous cell carcinoma group (female 16.67%, male 83.33%), and no statistically significant difference in the adenocarcinoma group (female 45.21%, male 54.79%) (Fig. 3c). Among patients with adenocarcinoma, there was no difference in TP53 positivity rate in males (27.38%) and females (23.81%), and the EGFR mutation was significantly higher in females (37.3%) than in males (26.19%) (Fig. 3d).

Mutation profiles of EGFR of our cohort
We found EGFR mutations in 94 samples of our cohort and we explored the mutational profiles of these samples. The integrated mutational landscape of EGFRmutant samples was illustrated in Fig. 4a

Discussion
In our study, 294 Chinese NSCLC patients from Shanxi Cancer Hospital were recruited and tested by commercially available NGS-based ctDNA assay, and 100% of them were successful in variant calling. NGS is an attractive platform to serve in clinical practice in a cost-effective and minimally invasive manner. To evaluate the analytical performance of our NGS-based ctDNA assay, we validated a set of detected and undetected mutations of ctDNA using ddPCR, which was regarded as the gold standard. We reported the concordance rate between NGS and ddPCR was 99.8% (418/419) for SNVs and 96.7% (146/151) for InDels. A high consistency between ctDNA assay and ddPCR arises. Overall, the most frequent altered genes detected in our cohort were: TP53 (32%), EGFR (31.97%), KRAS (6.46%), and PIK3CA (4.76%), which is broadly consistent with previous research findings (Chen et al. 2020;Schrock et al. 2019). Rosell et al. reported exon 19 deletion and point mutation L858R were the most common EGFR mutations, followed by L861Q, G719X and S768I, which were predictive of treatment benefit from EGFR-TKIs' therapy (Rosell et al. 2009). Similar results were found in our study. We found that the majority of KRAS mutations were in codon 12, and G12C was the most common mutation type, which was similar in previous studies (Shen et al. 2021). NSCLC has two major histological types: adenocarcinoma and squamous cell carcinoma, accounting for nearly 50% and 30%, respectively (Kim et al. 2013). However, 88.78% (261/294) were diagnosed with adenocarcinoma and 6.12% (18/294) with squamous cell carcinoma across the patients recruited in this study. Besides, the cohort included 126 females and 168 males. Interestingly, there is a statistically significant difference in the ratio of males and females in the squamous cell carcinoma group (female 16.67%, male 83.33%, p < 0.005). In this study, we used a uniform strategy to compare mutations in NSCLC across genders and major pathologies. We found the detection rate of EGFR was higher in female and patients with adenocarcinoma and conversely, patients in male and patients with squamous cell carcinoma had a higher detected rate of TP53 gene mutation. However, there was no difference in TP53  Exon 19 deletion and the point mutation L858R in exon 21 were the most frequent EGFR mutations in NSCLC in our study, which has been revealed by previous studies to be important in predicting the therapeutic effect of EGFR-TKIs ). However, Hong et al. reported that concomitant mutations could indicate inferior treatment outcome and L858R had a significantly higher incidence of concomitant mutation than 19del (Hong et al. 2018). In previous findings, longer survival and better responsiveness to EGFR-TKI therapy were observed in the subgroup with 19del/TP53 concomitant mutations compared to the patients with L858R/TP53 concomitant mutations (Hou et al. 2019). T790M is a widely discussed signature mechanism of primary resistance to EGFR-TKIs. It has been shown that more than half of patients treated with EGFR-TKIs were T790M positive, which are the most common alteration leading to acquired drug resistance . Interestingly, we have detected a higher proportion of L858R/TP53 concomitant mutations than the 19del/TP53 concomitant mutations, yet the proportion of 19del/T790M and L858R/T790M concomitant mutations was approximately the same. Our findings might give a clue the varied sensitivity of 19del and L858R to EGFR-TKIs.
There are some limitations in this study. A limited number of patients were enrolled in the retrospective study, and most patients lacked treatment information. It would be interesting to include more patient treatment information to explore the potential utility of ctDNA in patients with NSCLC in China, such as the relationship between the frequency of EGFR-sensitizing mutations and treatment effect. Unfortunately, we were only able to get sufficient treatment information for 13 patients and were unable to reach any firm conclusions based on the limited information. Only 20 genes were detected, and it was not possible to decipher the comprehensive molecular map in the ctDNA of Chinese NSCLC patients. Further studies based on extended gene and much more patient treatment information are needed to decipher the entire molecular profile in the ctDNA of Chinese NSCLC patients and to evaluate the clinical utility of ctDNA in clinical practice.
In conclusion, NGS-based ctDNA analysis can successfully and accurately reveal genomic alterations in the majority of NSCLC patients. We find a high agreement between the results of ddPCR and NGS. In our study, the most frequent altered gene was TP53, followed by EGFR, KRAS, PIK3CA and MET. Mutational profiles were associated with gender and pathology, with EGFR mutations significantly associated with female and adenocarcinoma. Exon 19 deletion and G12C were the most common mutation points in EGFR and KRAS, respectively. Our study also demonstrates that concomitant mutations are widespread in both 19del and L858R. The co-occurrence rate of somatic T790M in 19del was similar to that of L858R, yet the prevalence of TP53 in L858R was significantly higher than in 19del. Our findings suggest that the widespread use of ctDNA analysis may greatly benefit those patients whoare unable to perform invasive procedures and provide effective guidance for their medication management.