Since the dawn of precision medicine in oncology, tumour tissue biopsy has been the gold standard of genomic alteration profiling for solid malignancies. However, the acquisition and applications of tissue biopsy are not always straightforward due to several reasons, such as the invasive nature of the procedure, difficult-to-access anatomical location, and tumour heterogeneity. In recent years, liquid biopsy has gained quite a popularity in clinical applications, but are we ready to completely replace tissue biopsy with liquid biopsy?
With the advent and advancement of next-generation sequencing (NGS) technology, we have reached a point where the use of liquid biopsy – a non-invasive alternative to tumour biopsy, has increasingly gained popularity in clinical applications. Although tumour tissue biopsy is currently the gold standard of genomic alterations profiling for solid malignancies, its acquisition and applications are often temporally and spatially limited by technically challenging anatomy, complications associated with the surgical procedure involved, and sampling errors due to tumour heterogeneity.
Tumour biopsy provides a snapshot of the tumour status at a given time and location, but it may not reflect inter- and intra-tumour molecular heterogeneity known to exist in solid malignancies. Owing to these limitations, liquid biopsy has emerged as a useful tool for real-time monitoring of tumour biology and treatment selection in cancer patients from whom a tissue biopsy is unobtainable.
In this article, we will address the following pertinent questions and discuss the practical aspects of liquid biopsy in clinical applications:
- What are the advantages and limitations associated with liquid biopsy testing?
- Who may benefit from liquid biopsy testing?
- How can liquid biopsy be applied in cancer management?
- What considerations need to be taken when interpreting a liquid biopsy test result?
What is Liquid Biopsy?
In precision oncology, liquid biopsy refers to the collection of peripheral blood for analysis of circulating tumour-derived nucleic acids (ctDNA and ctRNA) or cells (CTCs). In the context of this article, only ctDNA will be discussed in details. CtDNA are fragmented DNA (~166 bp) passively released into the bloodstream by apoptotic/necrotic tumour cells or through active secretion by living tumour cells.1 It should not be confused with cell-free DNA (cfDNA), which is a broader term used to describe DNA fragments that are freely circulating in the blood, including those from normal cells.
CtDNA only constitutes a relatively small fraction of the total cfDNA pool and its quantity in the bloodstream varies among individuals, depending on the type of tumour, its location, disease burden, and cancer stage.2,3 The half-life of ctDNA in the bloodstream ranges from 16 minutes to 2.5 hours.3 The ctDNA fraction is usually less than 10 percent even at late stages.4,5 In solid malignancies, the proportion of patients with detectable ctDNA ranges from about 50 percent in non-metastatic disease to almost 90 percent in metastatic disease.6 After curative resection, the post-operative detection of ctDNA ranges from 10 to 15 percent of patients with stage II disease to approximately 50 percent in those with stage IV disease.7
Genomic alterations identified through a liquid biopsy may originate from ctDNA fragments, germline alterations, or non-tumour somatic alterations, such as clonal haematopoiesis of indeterminate potential, which may give rise to false-positive results.8 According to the latest National Comprehensive Cancer Network (NCCN) guideline, ctDNA testing generally has a high specificity, but a significantly compromised sensitivity has been reported with up to 30 percent false-negative rate.9 This notion is in line with the observation that about 22 percent of genomic alterations are detectable in the tumour only but not in the ctDNA (Figure 1).10 In view of these findings, NCCN guideline recommends that ctDNA should not be used in lieu of a histologic tissue diagnosis.
Advantages and Limitations of Liquid Biopsy
Liquid biopsies allow frequent and serial samplings over time to provide better resolution to tumour behaviours and treatment response. In one study colorectal cancer (CRC) patients who were later radiographically confirmed to be good responders to treatment had a drop of more than 90 percent of the ctDNA levels after the first 2 weeks of treatment.11 This information would allow clinicians to stratify patients based on the risk of recurrence after curative resection. In another study, breast cancer patients with detectable ctDNA after resection had a 25-fold higher hazard risk of recurrence.12 Liquid biopsy also sheds light on tumour heterogeneity since it contains ctDNA from various sources.
Important disadvantages of liquid biopsy include the need for an initial histologic diagnosis to be determined by tissue biopsy as it is unable to pinpoint the exact source of the ctDNA. Laboratories performing these assays need to be mindful of the appropriate test utilization and the potential for “over interpretation” in the clinical context. The low variant frequency within the peripheral blood may lead to higher false-negative rates and require significantly greater technical efforts and expertise to derive reliable interpretations of the test results.
Treatment Selection Using Liquid Biopsy
Liquid biopsy tests using a larger NGS panel – typically more than 50 genes – may represent an attractive means of genetic alterations profiling for treatment selection. However, a higher cost may limit the frequency of its use in the continuum of care, especially in the context of disease monitoring. Furthermore, a lower sequencing depth is often an inevitable trade-off for a broader gene coverage in order to maintain price competitiveness of a given panel.
One should also be cautioned that a negative liquid biopsy test result for any given variant does not preclude the presence of this variant in tumour tissue. If feasible, patients who are tested negative using a liquid biopsy should be reflexed to routine formalin-fixed paraffin-embedded (FFPE) tissue testing, so that they would not be erroneously excluded from potentially effective treatment options.
To date, the U.S. Food and Drug Administration (FDA) has approved a total 5 targeted drugs for solid malignancies in patients tested positive for the specific genetic biomarkers through a liquid biopsy (Table 1). Notably, genomic findings beyond the indicated genes which are also covered by the approved companion diagnostic panels are not prescriptive or conclusive for labelled use of other targeted therapies. In addition, clinical data for T790M plasma-positive patients are still limited. Therefore, liquid biopsy testing for treatment selection is most appropriate in patients from whom a tissue biopsy cannot be obtained.
What Contributes to a “Negative” ctDNA Test Result?
- Mutations occurred in genes which are not covered by the panel used
- CtDNA level falls below the panel detection limit due to disease status, tumour types or interventions (e.g., surgery or therapies)
- Inadequate ctDNA fraction and/or magnitude of copy number changes (CNVs) to infer a positive detection of CNVs due to the dilution effects of cfDNA from normal cells
- Genomic alterations which are unique to tumour tissue and not found in ctDNA
Who Would Benefit from ctDNA Testing?
Patients with solid tumours:
- when detection of post-operative minimal residual disease (MRD) is intended;
- when evaluation of treatment response is intended;
- who may be at risk of cancer recurrence;
- who progress on current treatment and drug resistance is suspected;
- who are medically unfit or contraindicated for invasive tissue biopsy.
Timing is Key in Liquid Biopsy
- A baseline profile must be established using the same panel before any intended interventions so that meaningful comparisons can be made
- Liquid biopsy should not be performed immediately after surgery or while patients are on active treatment due to an abnormally high plasma concentration of cfDNA released from the dying or dead cells that could confound the result
- If evaluation of MRD or treatment response is intended, allow a wash-out period of at least 2 weeks13 for the effects of surgery or therapeutic agents to subside and for the ctDNA level to stabilize before a liquid biopsy is obtained
Clinical Applications of ctDNA for Dynamic Monitoring of Tumour Mutations
- Early cancer detection (the idea is currently being evaluated in proof-of-concept studies; hence it will not be discussed herein)
- MRD detection
- Early relapse detection
- Treatment response evaluation
- Tracking clonal evolution and detecting drug resistance mutations
MRD Detection and Early Relapse Detection
In a prospective cohort study of 230 patients with stage II resected colon cancer, 1046 plasma samples were analysed by NGS assays to evaluate the ability of ctDNA to detect MRD and to identify patients at risk of recurrence.14 In patients without adjuvant chemotherapy, ctDNA was detected post-operatively in 14 of 178 (7.9%) patients, 11 of whom (79%) had recurred at a median follow-up of 27 months. While in the ctDNA negative group, recurrence occurred in only 16 of 164 (9.8 %) patients [HR: 18, 95% CI: 7.9 to 40, P<0.001]. In patients treated with adjuvant chemotherapy, the presence of ctDNA after completion of chemotherapy was also associated with an inferior recurrence-free survival [HR: 11, 95% CI: 1.8 to 68, P=0.001]. CtDNA detection after colon cancer resection provides direct evidence of MRD and identifies patients at very high risk of recurrence (Figure 3).15
Treatment Response Evaluation and Resistance Mutations Detection
The clinical relevance of ctDNA analysis for monitoring therapeutic response has been reported in patients with metastatic CRC.16,17 High basal ctDNA levels were associated with a short overall survival, and ctDNA assessment could be employed as an early surrogate marker of treatment response.18
Early changes in ctDNA levels during chemotherapy with molecular targeted drugs have been demonstrated to predict the treatment outcomes.11,13 Similarly, liquid biopsies have also been used to identify mechanisms of resistance to EGFR inhibitor therapy in patients with metastatic CRC.19 Notably, the emergence of resistant KRAS-mutated clones could be detected up to several months before radiological evidence of disease progression, indicating that ctDNA analysis through liquid biopsy is a more sensitive method for the early detection of relapse (Figure 4).20
Considerations in the Use of ctDNA
CtDNA constitutes a relatively small fraction of the pool of fragmented cfDNA circulating in the plasma, with the bulk consisting of normal DNA. Therefore, one should bear in mind that ctDNA copy number (CN) analyses reveal only relative CN changes, which neither allow the determination of the ploidy level of tumour cells, nor the exact prediction of the absolute CN due to the dilution effects with cfDNA from normal cells.
Reliable CN analyses require a relatively high allele frequency of ctDNA estimated to be at least 5 to 10 percent of plasma cfDNA.21 Hence, a low ctDNA fraction (e.g., during early disease stages or in non-shedding tumours) may potentially lead to a false-negative result for CNVs detection. It is imperative that clinicians are aware of these limitations, especially when it is utilized for treatment selection.
The strength of a ctDNA-based test lies in its ability to characterize the exact gene fusion breakpoints together with other genetic alterations, such as CNVs, single nucleotide variants, and small insertions and deletions. However, it is impossible to determine whether any given fusion event is expressed or not at the DNA level. Moreover, the detection of some fusion events involves intronic regions, which can be extremely large with repetitive sequences (i.e., NTRK2, NTRK3), thus hampering the sequencing efficacy and diagnostic accuracy.22
Despite being less stable than DNA, an RNA-based approach has several advantages:23
- It only identifies transcriptionally expressed fusion genes
- It distinguishes splicing isoforms
- It can quantify fusion transcripts
- It allows a contemporary analysis of exon skipping events
- Sequencing is not affected by intronic regions
The Future for Liquid Biopsy
Based on the information presented, it is clear that liquid biopsy simply cannot replace tissue biopsy in the settings of diagnosis and treatment selection. Well, at least not anytime soon. Nevertheless, liquid biopsy represents a useful tool for real-time monitoring of tumour biology. It is undeniable that liquid biopsy has its unique strengths and limitations which certainly warrant further investigation and validation in future larger prospective trials. Such efforts are particularly evident in the ongoing research of blood tumour mutational burden evaluation through liquid biopsy.
In view of the lack of standards for analytical performance characteristics of ctDNA tests, Friends of Cancer Research has recently launched a pilot project called “ctMoniTR”. This project aims to harmonize the use of ctDNA to monitor treatment response in cancer patients and test the feasibility of data comparison from different clinical studies. This project is broken out into two steps: Step 1 will use previously collected data from a subset of lung cancer trials to study the feasibility of ctDNA as a monitoring tool; While Step 2 will prospectively investigate the ability of ctDNA to detect early tumour response to treatment in clinical trials for different cancer types and treatments. Final results will be announced in 2022.