Potential of circulating tumour DNA as a biomarker in liquid biopsy

Abstract

Modern oncology has been influenced and transformed by precision and personalized medicine, especially with the introduction of personalized therapy. The ever-increasing knowledge of the genome contributes to the advancement of precision and personalized medicine. Liquid biopsy (LB) is a diagnostic approach with great potential for use in personalized medicine. LB is the isolation and analysis of tumour-derived components such as circulating tumour cells, circulating tumour DNA, circulating tumour RNA and tumour extracellular vesicles. The continuous development of new techniques in the field of LB holds promise for its widespread use. A challenge for scientists is the standardization of these techniques, allowing their application in clinical practice. This review presents the biological characteristics of circulating tumour DNA, popular methods for its detection and analysis, and its potential utility as a biomarker in liquid biopsy.

Keywords:

• Liquid biopsy
• circulating tumour DNA
• personalized medicine

Introduction

Precision and personalized medicine have become an integral part of oncology in the last decade. The development of ‘omics’ technologies has enriched the knowledge of genetic-molecular mechanisms regarding tumorigenesis. Molecular genetic analyses contribute to the implementation of a personal approach in the diagnosis, treatment and monitoring of the effect of therapy in oncology patients.

Tissue biopsy is a well-established and widely used method in clinical practice; it is the golden standard, by which a tissue sample is taken for genetic analysis to determine the type of tumour. Despite its widespread use, performing the tissue biopsy is a laborious process that requires specialized facilities and carries the risk of complications such as bleeding and spread of the tumour process [1].

Tissue samples may not always provide sufficient material or represent the heterogeneous nature of the tumour, thereby limiting their utility. Due to these disadvantages, tissue biopsy is unsuitable for mass cancer screening, monitoring the effect of therapy and detection of minimal residual disease [2].

Liquid biopsy (LB) is an innovative approach that offers diagnostic, prognostic and therapeutic opportunities by detecting, isolating and analysing various components of the tumour circulome [3]. The tumour circulome includes circulating tumour cells, circulating tumour nucleic acids such as DNA or RNA, which are either released from the cells or contained in exosomes, microvesicles and platelets. Body fluids such as blood, urine, cerebrospinal fluid, saliva, among others, serve as potential sources of tumour biomarkers for LB. This technique has the potential to overcome the limitations of tissue biopsy, making it a viable alternative.

The concept of liquid biopsy (LB) was initially proposed by Pantel and Alix-Panabières [4] as a means of exploring circulating tumour cells (CTCs) in blood samples. Alongside CTCs, several other biomarkers have been identified as promising candidates for use in LB, including cell-free DNA (cfDNA), with a particular emphasis on circulating tumour DNA (ctDNA), extracellular vesicles (EVs), circulating tumour RNA (ctRNA) and tumour-educated platelets (TEP) [5]. Previous reviews have focused on the use of liquid biopsy as a diagnostic and follow-up tool in cancer patients [6], others have focused on biomarker selection for liquid biopsy in lung carcinoma patients [7] or the use of ctDNA-based liquid biopsy in breast cancer patients [8]. This review aims to present a comprehensive analysis of the biological characteristics of circulating tumour DNA, the most used methods for its isolation, the different types of technologies and methods for subsequent analysis, and its potential utility as a biomarker for liquid biopsy in cancer patients.

Biological characteristics of cell-free DNA and circulating tumour DNA

CfDNA are DNA fragments that are either passively shed from necrotic and apoptotic cells or actively shed from living cells. In healthy individuals, cfDNA is rapidly phagocytosed by macrophages. In cases where the ability of macrophages to phagocytose is exhausted, nucleosome components are released mainly into the blood and other body fluids [9]. The average length of cfDNA in the blood of healthy people is 70–200 bp with concentrations in the range 0.3–10 ng/mL and a half-life between 16 min and 2.5 h [10]. An increased amount of cfDNA in the body is observed in various physiological and pathophysiological processes such as coagulation, immune reactions, cellular ageing, inflammatory processes, physical exertion and oncological diseases [2,11].

Cell-free ctDNA is a fraction of cfDNA originating from tumour cells [12]. The amount of ctDNA depends on the type and stage of the tumour. It can vary from 0.15% to over 10% of the total amount of cfDNA. The length of the ctDNA fragments ranged between 140–170 bp [13]. The liberation of ctDNA is known to occur via active and passive mechanisms. Passive release of ctDNA occurs upon cell apoptosis or necrosis. Active release is observed when ctDNA is released from living cells, which is directly related to the formation of nearby and distant metastases [14].

Isolation of ctDNA

CtDNA is contained in body fluids such as blood, cerebrospinal fluid, urine, and saliva. Of these, blood plasma is the preferred source for ctDNA isolation due to the absence of blood cells such as leukocytes, which can cause dilution of ctDNA levels. It is recommended to perform centrifugation of blood plasma within 4–5 h after sample collection to avoid degradation of ctDNA. While blood serum can also be used as a source of ctDNA, it is less preferred due to the potential for cell lysis during coagulation, leading to altered ctDNA concentrations [15].

Most ctDNA isolation methods follow similar principles. Blood samples are collected in tubes containing ethylenediaminetetraacetic acid (EDTA), which is an anticoagulant and inhibits blood DNаse activity [16].

It is recommended that blood samples collected in EDTA tubes should be centrifuged within 4–5 h to prevent contamination of the sample [17]. To ensure proper preservation of blood samples for cell-free DNA analysis, suitable containers containing preservatives are recommended. These include the Cell-Free DNA BCT by Streck, cf-DNA/cf-RNA preservation tubes by Norgen, Cell-Free DNA Collection Tube by Roche, DNA PAXgene Blood ccfDNA Tubes by Qiagen and others, which prevent the release and entry of nuclear DNA into plasma for several days. Improper storage can affect the volume of plasma available, resulting in an overall lower yield of ctDNA from the samples. To avoid ctDNA degradation, plasma should be stored at −80 °C, but prolonged storage (> 3 years) may result in subsequent reduced ctDNA yield [18].

According to what principle they are based on, ctDNA isolation methods can be divided into several groups, some of the most commonly used ones are:

Isolation methods based on a silica membrane spin column (silica)

The essence of the method is a membrane surface containing silica, on which nucleic acids bind under the influence of high concentrations of chaotropic salts and are separated at low concentrations of these salts. The method ensures a high yield and purity of the ctDNA samples. A disadvantage of this type of isolation is the complexity and laboriousness caused by the need for a high-speed centrifuge multiple reagents and consumables, and the partial loss of DNA fragments that are smaller than 150 bp [19].

Phenol chloroform isoamyl-based methods for ctDNA isolation

They are based on the denaturation of proteins by phenol. The advantage of these methods is that they use cheap and accessible reagents and are characterized by a high yield of ctDNA. A disadvantage is the high toxicity of the reagents to those who work with them; in addition, the isolated ctDNA is of low purity and yield quality, which prevents further analysis [15].

Methods for the isolation of ctDNA based on the magnetic enrichment of ctDNA

These methods are based on the use of magnetic beads with a positive charge that binds to the negatively charged phosphate backbone of DNA. This approach allows for the effective recovery of short ctDNA fragments and preserves the integrity of ctDNA due to the minimal effect of magnetism on metalation. The use of magnetic beads has several advantages, including easy automation, minimal sample loss and high purity and yield of the ctDNA sample [20].

To date, no scientific consensus has been reached on a standardized ctDNA isolation method. Therefore, it is important to carefully consider and select an appropriate isolation method that is in accordance with the subsequent ctDNA analysis method.

Analysis of isolated ctDNA

BEAMing is a method combining flow cytometry and PCR polymerase chain reaction. It involves performing single-molecule PCR reactions on magnetic beads in a water-oil emulsion. The BEAMing method consists of four main components, including magnetic beads, emulsion, amplification, and magnetism, with the process taking place in six main steps. During the method, the magnetic beads bind to DNA molecules specifically, and fluorescent labelling flow cytometry is used to assess the variation within the initial population of DNA molecules. Each emulsion droplet produces a PCR product that is attached to microgel beads, and after the PCR reaction is complete, the product is easily separated by magnetism, with the free DNA absorbed by magnetic beads [21]. BEAMing is highly sensitive, suitable for screening specific mutations, quantifying and identifying rare mutations, transcript variations, and gene sequences, and is relatively inexpensive and easy to perform. However, as a disadvantage, it is only suitable for detecting known mutations [22].

Droplet digital PCR (ddPCR) is based on a water-in-oil emulsion. The droplets are formed in a water-in-oil emulsion, in which the sample DNA molecules are separated into 20,000 nanolitre-sized droplets and PCR amplification is performed in each droplet individually. After the PCR reaction, each droplet is analysed to determine the fraction of PCR-positive droplets in the original sample. This allows the concentration of the target DNA template in the original sample to be determined [23]. The advantages of the method are high sensitivity, speed and relative economy. The method is suitable for targeted screening, monitoring the effect of therapy and minimal residual disease and allows absolute quantification of the target DNA sequence. As a disadvantage of the method, it is possible to analyse only a limited number of variants [18].

Real-time PCR (RT-PCR) is a method in which the DNA amplification is monitored in real time using a fluorescent signal that correlates with the number of amplified DNA molecules. The method involves several stages, including a linear phase at the beginning of the PCR reaction with a basic signal, an early exponential phase where the signal rises above baseline, a linear exponential phase where optimal amplification occurs with PCR products doubling each cycle, and a plateau phase where the signal no longer increases. Real-time determination of DNA amplification helps to avoid the risk of laboratory contamination during the reading stage. RT-PCR is a suitable method for both qualitative and quantitative analysis, transcript assessment, and detection of single nucleotide polymorphisms. A disadvantage of the method is its limited ability for multiplexing and that it detects only known mutations.

Next Generation Sequencing (NGS) is a highly versatile method that can detect a wide range of genetic variations, including single nucleotide substitutions, small insertions and deletions, inversions, duplications, and translocations. This technology can be applied to sequence entire genomes, exomes or specific panels of target genes [24]. NGS offers several advantages over traditional sequencing methods, such as higher throughput, greater sensitivity, more efficient use of DNA samples, and wider range of mutation detection. Although different NGS platforms may use different technologies, the general workflow consists of several stages, including library preparation, sequencing, and data analysis. In the library preparation stage, fragments of the DNA of interest are linked to adapter sequences During sequencing, each fragment is attached to a surface via the adapters and amplified by PCR to generate clusters that are then read in a single sequencing reaction. Finally, the obtained data are analysed to generate a comprehensive genetic profile of the sample [25].

Targeted Sequencing (TS) using NGS platforms has emerged as a rapid, accurate and cost-effective approach for detecting mutations in specific genes or gene targets. Compared to conventional sequencing techniques, TS methods offer greater accuracy and deeper coverage of the obtained data, resulting in more efficient mutation detection. NGS platforms commonly used for targeted sequencing include TAm-Seq (Tagged Amplicon Deep Sequencing), CAPP-Seq (CAncer Personalized Profiling by deep Sequencing), Safe-Seq (Safe Sequencing System) and AmpliSeq [26].

TAm-Seq, for instance, is capable of detecting and quantifying tumour mutations in a gene panel that includes both tumour hotspots and entire coding regions of selected genes [27]. This method enables parallel amplification of multiple regions and entire genes using a two-stage amplification approach. The first step is a pre-amplification with a pool of target-specific primer pairs that represents all alleles in the template material. In the next step, selective amplification of the desired gene regions is carried out through multiple reactions to exclude non-specific products. Finally, adapters and sample-specific barcodes are attached to the resulting amplicons through another PCR reaction [16].

Cancer Personalized Profiling by deep Sequencing (CAPP-Seq) is an additional next-generation sequencing (NGS) methodology. This technique utilizes biotinylated oligonucleotide selector probes that are designed to specifically target DNA sequences [28].

A specially designed NGS method that is more sensitive to rare mutations and reduces rate errors is Safe-SeqS. The essence of the method is based on the presence of a unique identifier (UID) to each sample molecule. A directly sequenced UID family is generated and used to amplify each sample molecule. Molecules with the same DNA sequence are considered to have the same UID. In this way, it becomes possible to identify ‘supermutants’, meaning that the UID family with approximately 95% of its members have the same mutation. The reported error rate in this case is 1.4 × 10⁻⁵ [29].

Targeted next-generation sequencing (NGS) technologies are well-suited for the rapid detection of specific mutations in each study, providing high levels of accuracy in mutation detection and sequencing depth. These technologies have the advantage of significantly reducing the detection process time and are particularly appropriate for analysing large numbers of samples, thereby effectively reducing sequencing costs.

Next-generation sequencing (NGS) technologies that employ a non-targeted approach for circulating tumour DNA (ctDNA) analysis include whole exome sequencing (WES) and whole genome sequencing (WGS). These techniques offer the significant advantage of enabling the identification of genetic changes across the entire exome or genome, providing comprehensive and detailed information regarding the mutational landscape of a given tumour.

Whole exome sequencing (WES). The WES method has the potential to detect activated somatic mutations by exome sequencing. WES could detect unknown genome-wide mutations leading to defective gene products, as well as detection of pathogenic and likely pathogenic variants in the protein-coding portion of DNA. A disadvantage of the WES method is that it does not detect mutations in the promoter and regulatory regions of genes and is limited to the coding and splice sites of genes [30].

Whole genome sequencing (WGS) is an innovative method for analysing whole genomes. Whole genome sequencing is suitable for identification of non-coding and structural variations. The advantage of the WGS method is that it performs whole genome analysis, it has wide use without personalization. The disadvantage of the method is the limited depth of sequencing, the high cost of the conducted research, the high complexity and low sensitivity [31].

Non-targeted approaches for ctDNA analysis are associated with certain disadvantages that can limit their clinical utility. One of the primary challenges is the significant amount of data generated by these methods, which can result in slow processing times and increased complexity in the diagnostic workflow. Additionally, non-targeted approaches may require a high starting concentration of ctDNA to achieve sufficient sequencing coverage, which can be a limiting factor in the detection of low-frequency mutations. Finally, the sensitivity of these techniques may be relatively low, which can contribute to longer turnaround times and a greater likelihood of false negatives.

Analysis of the methylation status of ctDNA is an emerging area of research that has shown significant promise as a potential biomarker for cancer screening. Methylated ctDNA can carry an epigenetic signature of the tissue of origin, which may be indicative of the presence of cancer. DNA methylation is a process by which a methyl group is added to CpG dinucleotides in regions of the genome known as CpG islands. The detection of methylated ctDNA can be achieved using a variety of techniques, including methylation-specific PCR (MS-PCR) and bisulphite treatment of DNA, which chemically modifies unmethylated cytosine into uracil. The methylation status of the converted DNA can then be analysed using a range of methods, such as PCR, next-generation sequencing (NGS), or MCTA-Seq (Methylated CpG Tandem Amplification and Sequencing) [32]. MCTA-Seq is a novel approach that involves a one-phase, three-step amplification of DNA fragments adjacent to CGCGCGG sequences, which are part of bisulphite-treated DNA. This technique is particularly useful for the detection of hypermethylated CpG islands and may have important implications for the development of new cancer screening methods [33].

Clinical application of ctDNA as a biomarker in liquid biopsy

CtDNA holds significant potential for LB applications as a diagnostic, prognostic, and monitoring biomarker for various cancers. The detection of ctDNA in body fluids has the advantage of being minimally invasive, less risky and less costly than conventional tissue biopsy. The use of ctDNA as a biomarker offers an opportunity for early detection of cancer, monitoring of tumour dynamics, and tracking response to therapy. Furthermore, ctDNA analysis provides the possibility of detecting genetic alterations in a non-invasive manner, and therefore has great potential for precision medicine [34].

Diagnosis of oncological diseases

LB based on ctDNA isolation and analysis was officially recognized as a diagnostic tool in 2016, but only for some specific mutations in certain tumours [9]. The quantitative and qualitative assessment of ctDNA has emerged as a promising approach for cancer screening and early diagnosis. Specifically, quantitative measurements of ctDNA in cancer patients have revealed elevated levels compared to those of healthy individuals [2]. Furthermore, the amount of ctDNA has been shown to correlate with disease severity and the presence of metastases, suggesting its potential as a prognostic marker.

Emmanuelle Gormally et al. [35] reported that clinically healthy individuals have detectable mutations in ctDNA up to two years before the onset and clinical diagnosis of cancer. Bettegowda et al. [36] found that ctDNA was present in 47% of stage 1 cancer patients and 80%–100% of stage 4 cancer patients. In clinically healthy individuals, ctDNA has been found to provide information on subclinical conditions as well as information on unknown neoplasms. In some of the clinically healthy patients in whom ctDNA-related mutations are found, cancer develops over time. Another group of healthy volunteers, with detected ctDNA mutations, did not develop cancer. This indicates that performing ctDNA-based mass screening in asymptomatic individuals carries the risk of overdiagnosis [37,38]. In order to routinely use ctDNA as a biomarker for cancer screening, it is necessary to detect and demonstrate mutations with high prognostic value and penetrance for oncological diseases [6].

In 2016, the United States Food and Drug Administration (FDA) granted official approval [FDA-2018-D-3380] for the Cobas EGFR Mutation Test v2, marking the first time such approval was granted. The test is designed for the detection of epidermal growth factor receptor (EGFR) mutations in the plasma of patients who have non-small cell lung cancer (NSCLC) [39]. Following this milestone approval, the FDA has since approved additional methods for LB that can detect specific gene mutations and rearrangements in patients with ovarian, lung, breast and metastatic prostate cancer [6].

A study conducted with CancerSEEK tests by Cohen et al. [40], including patients with a variety of proven non-metastatic tumours, found that on average 70% of all patients tested positive, the sensitivity of positive tests ranged between 69% and 98%, and the specificity was more of 99%

It has been shown that ctDNA analysis can detect a PIK3CA mutation, which identifies breast cancer with 95% accuracy, and lung cancer is identified with the detection of KRAS EGFR mutations with up to 97% accuracy. In conclusion, ctDNA analysis has the potential to detect mutations of interest and genetic heterogeneity [41].

Diagnosis of minimal residual disease (MRD). Detection of ctDNA after treatment is associated with a high risk of relapse. The presence of residual cancer cells, in cancer patients after surgical treatment and or radiation-chemotherapy, without clinical or radiological signs of metastases or residual tumour lesion is called minimal residual disease. MRD has the potential to lead to local recurrence and metastasis of the tumour process [42].

Imaging studies and conventional tests, such as blood protein tumour markers, have been shown to be incapable of detecting residual tumour cells. The identification of MRD is thus solely achievable using molecular genetic tests. MRD represents a significant prognostic indicator with respect to the prediction of both relapses and metastases. Given the importance of early detection in the context of improving patient outcomes, the identification of MRD is paramount for facilitating early intervention, and ultimately, enhancing the likelihood of survival for cancer patients [43].

The information obtained via analysis of ctDNA |represents a valuable resource for patient stratification and the assessment of recurrence risk, thereby supporting a personalized therapeutic approach. Furthermore, ctDNA analysis can contribute to the establishment of a comprehensive molecular profile for each tumour [43]. Notably, the identification of ctDNA after the conclusion of therapy or during the monitoring of ongoing therapy has been shown to be highly predictive of an elevated risk for recurrence in numerous cancer types, including early-stage breast cancer [44], colorectal cancer [45], lung [46] and bladder cancer [47].

Following surgical intervention, genetic alterations can be identified from resected tumour tissue, which can be utilized to generate an individualized plasma ctDNA panel. Ng and colleagues [48] undertook a study involving 44 patients afflicted with colorectal cancer. Within this study, patient-specific ctDNA assays were devised based on the multiplex detection of somatic mutations that were previously discovered from the patients’ primary tumours. They identified the existence of ctDNA in all patients prior to the operation, and only in those patients who suffered a relapse post operatively. In 11 out of 15 relapsed patients, ctDNA was found to be detectable prior to clinical or radiological indications of recurrence [49].

In addition to its prognostic potential for predicting relapse, the detection of MRD via ctDNA analysis presents another promising avenue for optimizing adjuvant and consolidation therapies. Specifically, the aim of this benefit is to enable the identification of MRD-positive patients who may benefit from additional therapeutic interventions, while possibly sparing MRD-negative patients from unnecessary adjuvant therapy [50,51].

The detection of MRD is a critical factor in determining the long-term survival of cancer patients. However, for this purpose, specially designed ctDNA assays are necessary, as conventional ctDNA analysis methods utilized for detecting solid tumours lack sensitivity for detecting MRD and are susceptible to producing erroneous results. Current analytical approaches demonstrate a low sensitivity of less than 50% for detecting MRD following treatment. Therefore, the development of high-sensitivity assays capable of identifying ultra-low concentrations of ctDNA is imperative for accurately predicting late relapse and reducing the incidence of false-negative samples [34].

Monitoring the effect of the therapy. On occasion, the administration of chemotherapy may prove ineffective, and associated with a concomitant escalation in toxic side effects with increasing dosages, primarily due to tumour resistance arising from the inherent heterogeneity of the malignancy. Given that tumour genome evolution during treatment cannot be adequately monitored via tissue biopsy, the use of tumour biopsies in conjunction with ctDNA analysis presents a valuable alternative [26].

The first-line therapy for patients with advanced NSCLC is epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKIs). Resistance is observed in a large proportion of patients treated with first- or second-generation EGFR-TKIs. In approximately 50% of resistance cases, the T790M mutation is detected. In patients with advanced EGFR-TKI-sensitizing mutant (EGFRm) or metastatic NSCLC treated with osimertinib, the absence of EGFRm in plasma was found to be associated with progression-free survival (PFS) compared with patients in whom plasma EGFRm was detected. These data demonstrate the prognostic potential of ctDNA analysis in EGFR-targeted therapy [52]. A study by Miteva-Marcheva et al. [53] found that liquid biopsy has potential in detecting germline mutations potentially associated with chemotherapy-induced toxicity in patients with colorectal and non-small cell lung cancer.

Notably, certain mutations in the KRAS, BRAF and PI3KCA genes have been implicated in the development of resistance to Regorafenib in patients with CRC. Furthermore, the demonstrable quantitative changes in ctDNA levels in CRC patients following chemotherapy underscore the potential utility of ctDNA as a biomarker for monitoring therapeutic efficacy [37].

In a study involving 53 CRC patients with previously established mutations and subsequent follow-up of the effect of ctDNA therapy, early changes in ctDNA during chemotherapy were shown to be associated with treatment response at the first and second doses of chemotherapy. In the study conducted, it was observed that the median fraction of ctDNA which was 16.24% dropped to 0.54% by the start of the second chemotherapy, but with no significant early change in carcinoembryonic antigen (CEA) levels [54].

To generate future guidelines, further studies are warranted to standardize procedures pertaining to the use of ctDNA as a biomarker in this context [12,55].

Conclusions

LB is an innovative method with enormous potential for use in precision and personalized medicine. This innovation is based on the detection and isolation of ctDNA, circulating tumour cells and exosomes. The most promising and reliable biomarker in liquid biopsy is ctDNA. The many studies in this direction will contribute to the application of LB based on ctDNA in clinical practice. LB has the potential to replace tissue biopsy in hard-to-reach tumour localization, repeat biopsies if necessary, and follow-up the disease over time. Applications of great clinical significance of LB are early cancer diagnosis and cancer screening; the monitoring of therapy in cancer patients with the aim of evaluating and personalizing therapy for each patient, as well as avoiding side effects of therapy; detection of MRD and metastases, which is essential for the survival of cancer patients. To introduce LB into routine practice for all oncological diseases, it is necessary to establish the standardization of preanalytical and analytical variables of existing LB ctDNA-based methods. It is also necessary to establish standard protocols for the isolation and standardized analysis of ctDNA involving all types of oncology, which should be standard for all laboratories conducting these tests. At low concentrations of ctDNA, it is necessary to use isolation methods enriching the samples and facilitating the subsequent analysis. The introduction of new technologies such as fragmentation pattern sequencing, methylation-based sequencing and ultra-sensitive mutation detection technologies will optimize the utility of LB. Despite the limitations still present, LB is a safe and reliable method for better and early diagnosis, monitoring the effect of therapy and the evolution of oncological diseases.

Author contributions

DD: conceptualization, writing—original draft preparation; HI: supervision, writing—editing. NM-M: writing—editing. VS: supervision, writing—reviewing and editing. All authors have read and approved the final version of the paper.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Data are available upon request.

Additional information

Funding

The authors received the support (financial, computational, logistic) from the project № BG-RRP-2.004-0007-С01 Strategic Research and Innovation Program for the Development of MU – PLOVDIV–(SRIPD-MUP), Creation of a network of research higher schools, National plan for recovery and sustainability, financed by the European Union – NextGenerationEU.