Circulating tumour DNA testing in metastatic breast cancer: Integration with tissue testing

Breast cancer biomarker profiling predominantly relies on tissue testing (surgical and/or biopsy samples). However, the field of liquid biopsy, particularly the analysis of circulating tumour DNA (ctDNA), has witnessed remarkable progress and continues to evolve rapidly. The incorporation of ctDNA‐based testing into clinical practice is creating new opportunities for patients with metastatic breast cancer (MBC). ctDNA offers advantages over conventional tissue analyses, as it reflects tumour heterogeneity and enables multiple serial biopsies in a minimally invasive manner. Thus, it serves as a valuable complement to standard tumour tissues and, in certain instances, may even present a potential alternative approach. In the context of MBC, ctDNA testing proves highly informative in the detection of disease progression, monitoring treatment response, assessing actionable biomarkers, and identifying mechanisms of resistance. Nevertheless, ctDNA does exhibit inherent limitations, including its generally low abundance, necessitating timely blood samplings and rigorous management of the pre‐analytical phase. The development of highly sensitive assays and robust bioinformatic tools has paved the way for reliable ctDNA analyses. The time has now come to establish how ctDNA and tissue analyses can be effectively integrated into the diagnostic workflow of MBC to provide patients with the most comprehensive and accurate profiling. In this manuscript, we comprehensively analyse the current methodologies employed in ctDNA analysis and explore the potential benefits arising from the integration of tissue and ctDNA testing for patients diagnosed with MBC.


| INTRODUC TI ON
The management of breast cancer is an ever-evolving landscape, driven by the introduction of actionable biomarkers and the continuous refinement of testing strategies. 1 This progress is particularly pronounced in the context of metastatic breast cancer (MBC), in which emerging biomarkers provide deeper insights into the disease landscape and tailor treatment approaches. 2 Biomarker assessment mainly relies on analyses that are carried out on tissue samples obtained through surgical procedures or biopsies. 3However, such analyses can be troubled by tumour heterogeneity, sampling limitations, and invasiveness. 4The notable breakthrough of liquid biopsy, especially with the detection of circulating tumour DNA (ctDNA; i.e., fragments of DNA derived from tumour cells) has provided an alternative approach to face the limitations related to traditional tumour samples. 5 this regard, liquid biopsy, which encompasses the analysis of different components found in biofluids, including DNA, RNA, circulating tumour cells, and extracellular vesicles, holds immense promise and is experiencing rapid expansion for patients with MBC. 6 Specifically, ctDNA analysis provides valuable insights into the genetic alterations and dynamics of tumours, offering several benefits in disease management. 7ctDNA testing demonstrates its utility in various aspects including early detection of disease progression, tracking treatment response, biomarker testing and monitoring, and identification of resistance mechanisms. 8Nevertheless, it is imperative to acknowledge certain limitations inherent to ctDNAbased liquid biopsy.Technical constraints, including the sensitivity and specificity of the currently available testing methods (i.e., polymerase chain reaction [PCR]-based and next-generation sequencing [NGS]-based) can result in false-negative findings, 9 whereas natural events such as clonal haematopoiesis of indeterminate potential (CHIP) have been linked to false-positive results. 10nsidering the logistic and technical challenges associated with tissue and ctDNA-based testing respectively, the development of a diagnostic workflow that will opt for the integration of both approaches is warranted. 11The current work provides a comprehensive overview of the methodologies currently utilized for ctDNA detection and highlights the growing relevance of this analysis in MBC.Moreover, we critically examine different scenarios that stand to gain remarkable advancements by integrating ctDNA and tissue testing.

| c tDNA DE TEC TION ME THODS
In cancer patients, the presence of ctDNA is typically observed at a relatively low proportion, ranging from 0.01% to 1.0% of the total cell-free DNA (cfDNA). 12The concentration of ctDNA is generally <1 ng/μL, and its quantity can vary based on factors such as stage, location, or vascularisation of the tumour. 13Consequently, there is a clear demand for highly sensitive techniques capable of detecting the genomic characteristics of limited ctDNA quantities.To address this need, various detection methods have been developed, primarily falling into two distinct categories: PCR-based and NGS-based methods (Figure 1).

| PCR-based techniques
Among the methods that have been developed so far, real-time quantitative PCR (qPCR); Beads, Emulsion, Amplifying, and Magnetics (BEAMing); and droplet-digital PCR (ddPCR) are the most prevalent, offering a simple and cost-effective approach for the detection and quantification of ctDNA. 14qPCR enables the selection of assays with different reference ranges for the identification of mutations in known genomic regions, providing results in a short turnaround time.However, qPCR-based assays can only detect a few hotspot mutations at a time and their limit of detection ranges between 1% and 5%. 15ddPCR and BEAMing have shown remarkable sensitivity (0.01% to 0.1%) in detecting somatic point mutations. 16The former requires a sequential two-step process to enable the precise analysis of DNA fragments.In the initial emulsion step, numerous lipid droplets are generated, facilitating the isolation of individual DNA fragments.The second step entails the execution of PCR within each encapsulated droplet.By employing primers that are uniquely labelled with distinct fluorophores, the identification of both mutated and unmutated copies is accomplished. 17Considering its technical features and practical suitability for routine implementation, ddPCR represents a valuable option in certain scenarios such as the identification of mutations occurring in the oestrogen receptor (ESR1) gene within the setting of oestrogen receptor-positive/human epidermal growth factor receptor 2-negative (ER+/HER2−) breast cancer. 18By contrast, BEAMing, which relies on the techniques of PCR and flow cytometry, 19 is characterized by a laborious workflow and thus its true implementation in the diagnostic workflow of MBC remains to be defined. 20

| NGS-based techniques
NGS assays present a unique opportunity to simultaneously analyse multiple molecular targets by multiplexing samples in a single run. 21However, the unique characteristics of ctDNA have driven the development of novel, more powerful NGS methods that have significantly enhanced the limit of detection through various technical adaptations.These include deep-sequencing coverage, molecular barcoding, and error-correction algorithms. 22Approaches that are often adopted include tagged amplicon deep sequencing, cancer personalized profiling by deep sequencing, and safe sequencing system. 23In these methodologies, barcoded cfDNA sequencing libraries are created, leading to the generation of multiple redundant sequences originating from the same barcoded molecule.By consolidating these cfDNA molecules, it becomes possible to differentiate between true mutations and sequencing errors.True mutations are expected to be present in multiple NGS reads containing the same barcodes. 20Tagged amplicon deep sequencing, introduced by Forshew et al, 24 demonstrates exceptional sensitivity (>97%) in detecting ctDNA mutations in plasma, even at very low allelic frequencies (~2%), whereas deep sequencing for cancer personalized profiling, which has been developed by Newman et al, 25 allows the detection of ctDNA mutant fractions as low as 0.02% while maintaining a high level of specificity (~95%).While the aforementioned techniques excel in their ability to target and detect specific genetic mutations with high analytical sensitivity, they present limitations in providing a holistic perspective of an entire tumour, unlike whole-exome sequencing (WES) or whole-genome sequencing (WGS) approaches.
However, it is important to note that WES and WGS have certain limitations that restrict their utility in specific cases.For instance, these methods tend to have lower overall sensitivity, typically around 5% to 10%.Additionally, they require higher concentrations of ctDNA, 5 which can be challenging to achieve, particularly in patients with low ctDNA load. 26Furthermore, WGS presents further challenges in terms of cost-effectiveness, prolonged turnaround times, and difficulties in interpreting results outside of specialized centres.These factors hinder the widespread adoption and practical implementation of WGS in routine analyses in the field of predictive pathology. 27deed, it is important to acknowledge the significant role that WES and WGS play in delivering a comprehensive and detailed analysis of the entire tumour genome.These approaches provide a broader perspective of genetic alterations, encompassing known and novel mutations, structural variations, and non-coding regions.While they have drawbacks, a balanced approach is necessary, considering the trade-off between the comprehensive coverage offered by WES/ WGS and the enhanced sensitivity and targeted analysis provided by the aforementioned techniques.This consideration implies that the sequencing approach should be tailored according to specific clinicopathological parameters.

| APPLIC ATIONS OF c tDNA ANALYS IS I N M E TA S TATI C B R E A S T C A N CE R
ctDNA can be detected in a great proportion of patients with MBC, in contrast to what usually occurs in early-stage disease. 28Of note,

Zhou et al demonstrated that approximately 86% of patients with
Stage IV/M1 breast cancer harboured tumour-derived mutations in their blood, whereas only 57.8% of patients with Stage I-III/M0 breast cancer exhibited such mutations. 29In the metastatic setting, the analysis of ctDNA may offer valuable biological information ranging from monitoring and predicting the progression of late-stage disease, to assessing mutations in actionable genes such as PIK3CA F I G U R E 1 Graphical representation of the diagnostic methods that are currently available for circulating tumour DNA (ctDNA) genotyping.Analytical methods that provide the greatest breadth or nucleotide coverage have lesser sensitivity and require higher fractions of ctDNA in overall cell-free DNA for accurate analysis.BEAMing, Beads, Emulsion, Amplifying, and Magnetics; CAPP-Seq, cancer personalised profiling by deep sequencing; ddPCR, droplet-digital PCR; NGS, next-generation sequencing; PCR, polymerase chain reaction; qPCR, real-time quantitative PCR; Safe-Seq, safe sequencing system; TAM-Seq, tagged amplicon deep sequencing; WES, whole-exome sequencing; WGS, whole genome-sequencing.and ESR1 (Figure 2).These pieces of information are pivotal for clinical decision-making.

| Testing for actionable biomarkers
The evaluation of specific genetic alterations in genes such as PIK3CA and ESR1 holds paramount significance in the context of MBC, playing a crucial role in guiding the careful selection of treatment protocols.Several studies have provided compelling evidence supporting the routine evaluation of such biomarkers, either in plasma and tissue samples or even in plasma alone, with the potential to enhance personalized medicine approaches and improve outcomes for patients with MBC.

| PIK3CA
Activating mutations in PIK3CA, the gene encoding the p110α subunit of PI3K, are prevalent in approximately 40% of breast cancers. 30ese mutations result in the dysregulation of the PI3K/AKT/mTOR signalling pathway, which plays a crucial role in cell growth, proliferation, and survival. 31In the SOLAR-1, a phase III randomized study, significant benefits were observed in patients with PIK3CAmutated hormone receptor-positive, HER2-negative (HR+/HER2−) advanced breast cancer who were treated with alpelisib (BYL719; Novartis Pharma AG), an α-selective PI3K inhibitor and degrader, in combination with fulvestrant.The study demonstrated improved progression-free survival (PFS) rates in the alpelisib plus fulvestrant group compared to the control group. 32In this trial, PIK3CA mutational status was tested both on the available tissue sample (primary tumour or metastasis) and plasma collected at the time of study entry.The results showed remarkable levels of specificity, with a negative agreement of 97% between liquid biopsy and conventional tissue biopsy (209 of 215); however, the sensitivity of ctDNA testing was comparatively lower, with a positive agreement of 55% (179 of 328). 33Based on these results, alpelisib received approval from both the U.S. Food and Drug Administration (FDA) and European Medicines Agency.The approval includes the use of a molecular test for PIK3CA that can be employed on tissue specimens and/or ctDNA. 34The FDA granted approval for the companion diagnostic test therascreen® PIK3CA RGQ PCR Kit, while the European Medicines Agency recommends the use of a validated test.It is important to exercise caution when selecting an appropriate assay, since methods with relatively low sensitivity and restricted reference range may lead to false-negative results.
In addition to these findings, several translational research studies investigated the agreement of PIK3CA status between matched tumour and plasma-derived ctDNA data.In one study involving a cohort of 46 patients with MBC, a concordance rate of 95% was observed. 35Another recent study published by Suppan et al, which included a total of 72 patients, reported a moderate agreement between tissue and plasma samples with an overall concordance rate of 72.2%. 36Both studies provided context for the observed discrepancies.While some of them were attributed to the timing of blood sampling, suggesting that ctDNA quantity varies over time, in other cases, the analysis of a primary tumour tissue sample did not accurately represent the mutational status of F I G U R E 2 Schematic representation of the current actionable biomarkers in metastatic breast cancer able to be tested with ctDNAbased analyses.The internal and external cycles refer to the type of diagnostic tests (i.e., CDx, CLIA, or LDT) and the level of validation (i.e., breast, pan-cancer, or other tumour type-specific) of each biomarker, respectively.Both levels of information are colour-coded on the basis of the legend at bottom right.bTMB, blood-based tumour mutational burden; CDx, companion diagnostic; CLIA, Clinical Laboratory Improvement Amendments-certified; ctDNA, circulating tumour DNA; LDT, laboratory developed test; MSI, microsatellite instability.
the metastasis, likely due to tumour evolution and intra-tumoural heterogeneity.

| ESR1
During the course of treatment with aromatase inhibitors, resistance phenomena may be observed in patients with MBC. 37One significant mechanism of resistance is the acquisition of mutations in ESR1, which occurs in around 30% to 40% of cases. 38This condition has been investigated in several studies which identified the presence of subclonal mutations in ESR1 after patients' exposure to endocrine therapy. 39,40The detection of these mutations has been made possible through ctDNA-based analyses.Of note, the BOLERO-2 trial specifically examined Y537S and D538G mutations in ESR1, demonstrating their association with reduced overall survival and PFS. 39Similarly, in the PALOMA-3 trial, treatment with fulvestrant alone and/or in combination with palbociclib led to the emergence of the ESR1 Y537S mutation, which confers resistance to treatment. 41In the former trial, ESR1 assessment was conducted using ddPCR, whereas the latter was based on a combination of ddPCR, targeted sequencing, and WES.This approach provided additional valuable information by revealing an increased detection of mutations at the end of treatment compared with initial samples.In addition to these insights, the randomized phase III EMERALD trial investigated the efficacy and safety of elacestrant, an oral selective oestrogen receptor degrader, versus endocrine monotherapy in patients with previously treated ER+/HER2− advanced breast cancer, including patients with ESR1-mutated tumours. 42Improved PFS was reported both in the overall population and in patients with ESR1 mutations in ctDNA.According to these results, the American Society of Clinical Oncology expert panel recommends routine testing for ESR1 mutations at disease recurrence or progression for patients with HR+/HER2− MBC. 43In conjunction with the approval of elacestrant, the FDA has also authorized the use of the NGSbased Guardant360 CDx assay as a companion diagnostic device.Furthermore, the PADA-1 trial represents the first prospective randomized trial to incorporate therapeutic decisions guided by ctDNA findings in HR+/HER2− advanced breast cancer.Notably, patients with newly detected ESR1 mutations in ctDNA demonstrated significantly prolonged PFS when treated with a combination of the selective oestrogen receptor modulator fulvestrant and the CDK4/6 inhibitor palbociclib, in comparison to cases undergoing traditional aromatase inhibition plus palbociclib therapy. 44,45These recommendations open new avenues in the incorporation of ctDNA-based testing in MBC management.Nevertheless, there is a growing interest regarding ESR1 fusions, and their role as an additional mechanism of resistance to endocrine therapy. 46,47Despite their rarity, these events are found more frequently in patients with aggressive tumours such as endocrine therapy-resistant MBC and Luminal B breast cancer. 48However, the evaluation of ESR1 fusions has not yet become part of the routine practice, and thus further data are requested.

| ERBB2 mutations
In a relatively small proportion of patients with MBC (2% to 4%), occurrence of ERBB2 mutations in ERBB2 non-amplified tumours has been observed. 49Neratinib, an irreversible inhibitor that binds HER2 and inhibits downstream signalling, has shown benefit when used in combination with fulvestrant in post-menopausal MBC patients with ERBB2 mutant, non-amplified tumours. 50Acquired ERRB2 mutations (e.g., L755S) have been associated with resistance to lapatinib but sensitivity to neratinib, highlighting the importance of serial plasma analyses for their detection in ctDNA. 51Indeed, ctDNA sequencing offers a non-invasive strategy to identify ERBB2 mutant cases, with high specificity and consistent sensitivity, and can help monitor disease progression through quantification of ERBB2 mutation allele frequency or identification of additional mutations. 52Additionally, in HER2-amplified breast cancer patients, the acquisition of the HER2 V777L mutation leads to trastuzumab resistance; this mutation as well can be detected by means of liquid biopsy methods. 51These findings strongly support the rationale for ctDNA-based testing of ERBB2 mutations in patients with MBC.

| NTRK fusions
NTRK1, NTRK2, and NTRK3 genes encode for the three neurotrophic receptor tyrosine kinase proteins NTRK1 (also known as TRKA), NTRK2 (also known as TRKB), and NTRK3 (also known as TRKC), respectively, which play crucial roles in the function of the nervous system. 53When fusions occur between the 3′ kinase domain of NTRK and the 5′ region of various other genes, it can result in the constant activation of TRK, leading to uncontrolled downstream signalling and the development of tumours. 54though rare, NTRK fusions are present in both secretory and non-secretory breast cancer subtypes as suggested by recently published data. 55Importantly, NTRK fusion-positive tumours can be treated with TRK small molecule inhibitors (e.g., larotrectinib and entrectinib) based on the tumour-agnostic approval that was granted by the FDA. 56The main technologies for NTRK fusion testing encompass DNA-and RNA-based NGS, qPCR, immunohistochemistry, and fluorescent in situ hybridisation, carried out on tissue samples.In 2019, a strategy for the most appropriate implementation according to samples availability was proposed by the European Society for Medical Oncology's Translational Research and Precision Medicine Working Group. 57However, recent lines of evidence suggest potential utilisation of liquid biopsy for gene fusions, including those occurring in NTRK genes. 58Pan-cancer studies reported a higher than 85% concordance between tissue and plasma in detecting NTRK fusions, suggesting the feasibility of this analysis through ctDNA testing. 59,60Nevertheless, in a series of patients with lung cancer, Hasegawa and colleagues investigated the analytical validity and clinical usefulness of fusion detection through circulating cell-free RNA (cfRNA) analysis showing its higher sensitivity compared with ctDNA. 61These results question whether ctDNA or cfRNA should be the preferred analyte to be used for the detection of such molecular alterations, suggesting that further investigations are required.

| Tumour mutational burden (TMB) and microsatellite instability (MSI)
With the introduction of immunotherapy, various biomarkers have been proposed to predict response to immune checkpoint inhibitors, including tumour mutational burden (TMB) and microsatellite instability (MSI), which assess the quantity of non-synonymous mutations in cancer and the status of microsatellites, respectively.Both these biomarkers are primarily tested on tissue samples; however, recent evidence suggests potential utility of blood-based analyses considering their non-invasive nature.The predictive value of bloodbased TMB (bTMB) has been extensively studied in non-small cell lung cancer patients both retrospectively (e.g., POPLAR 62 and MYS-TIC 63 trials) and prospectively (e.g., NEPTUNE, B-F1RST, and BFAST trials).However, the primary endpoints, including overall survival or PFS in the bTMB-high cohorts, were not consistently met, likely due to substantial variations in the predefined bTMB cut-off values across these studies (i.e., 16-20 mutations/Mb). 64In the context of breast cancer, a higher prevalence of high bTMB has been observed in patients with MBC. 65Nevertheless, before bTMB can be employed to predict immunotherapy response, further standardisation and harmonisation steps are necessary.These include adjustments for sequencing depth, coverage of specific regions of interest covered by NGS panels, and identification of insertions and deletions (indels) resulting from frameshift alterations, which are known to have higher immunogenicity than single nucleotide substitutions. 66I is a phenomenon that arises from inherent deficiencies in DNA mismatch repair and can be assessed through ctDNA analyses to predict response to immunotherapy. 67In breast cancer, MSI occurs in approximately 2% of cases. 68The main detection methods include PCR and NGS analyses conducted on tumour tissue specimens. 69Detecting MSI in liquid biopsies requires the integration of molecular barcoding with deep NGS and the use of advanced bioinformatic algorithms capable of reducing NGS-associated artefacts and detecting subtle alterations in microsatellite lengths. 70In this context, two studies conducted by Georgiadis et al and Willis et al have brought remarkable technological improvement.The former adopted a multifactorial error correction approach for precise calling of indels in cfDNA fragments along with a digital peak finding algorithm for quantification of MSI high alleles. 71In the latter study, the authors employed a digital sequencing error correction method to accurately define true indels at microsatellite loci. 67To date, FoundationOne Liquid CDx, an FDA-approved companion diagnostic ctDNA assay, is able to analyse bTMB and MSI but is generally troubled by relatively extended turnaround time. 72Thus, despite the promising advancements, the in-house implementation of the aforementioned sophisticated approaches into routine practice is still challenging and requires further development.

| Tracking response to treatment: Tumour burden and clonal heterogeneity
Serial plasma monitoring in patients with MBC offers clinicians a valuable opportunity to promptly assess the response to ongoing treatments.Changes in ctDNA levels demonstrate a strong correlation with tumour burden, making ctDNA a highly valuable noninvasive tool for monitoring tumour evolution, predicting treatment response, and determining prognosis. 73Compared with circulating tumour cells, ctDNA is more abundant; however, it is characterized by rapid clearance from circulation within a few hours. 74Notably, an increase in ctDNA levels has demonstrated the ability to anticipate disease progression several months earlier than conventional imaging techniques.It has also shown superior efficacy in assessing treatment response earlier than other established markers. 75Additional research has indicated that the dynamics of ctDNA during the early stages of treatment might also provide insights into the clonal composition of the tumour upon disease progression. 64A study conducted by Murtaza and colleagues analysed multiregional tumour biopsies and plasma samples collected from one patient with metastatic ER+/HER2− breast cancer, and showed that sequential alterations in the levels of subclonal private mutations detected in the circulation were directly correlated with diverse treatment responses observed across metastatic sites. 76Notably, in the randomized phase III PALOMA-3 study, where a combination of CDK4/6 inhibitor and fulvestrant was employed in advanced breast cancer patients, ctDNA analyses on plasma samples suggested that early ctDNA dynamics can reveal distinct responses of various tumour subclones to the treatment. 77Identifying these variants and understanding their dynamics could have a profound impact on the treatment trajectory and potential response in the event of disease relapse.However, it is essential to acknowledge that despite these promising advancements, the utility of such techniques is not yet fully realized, as there have been documented cases of inconsistent responses between ctDNA and imaging scans. 78This discrepancy is likely attributed to the tumour heterogeneity and the subclonal nature of MBC.To make this surveillance method relevant, further investigations and development are required to address assay sensitivity, subclonal dynamics, and the emergence of resistance mutations.

| Expanding the molecular profiling
Tumour tissues are considered the gold standard biospecimens to carry out molecular analyses in breast cancer. 79However, there are certain conditions where their intrinsic limitations hinder appropriate molecular profiling. 80Of note, analyses conducted on tissue samples portray tumours' state at a specific point of the disease course without being able to capture intra-tumoural heterogeneity.
Comprehensive molecular profiling of multiregional metastases of breast cancer patients have demonstrated intra-tumoural and intertumoural differences which may be underestimated when relying on single tumour biopsies. 76,81,82Thus, integrating liquid biopsy into the diagnostic workflow could provide relevant information in addition to initial tissue genotyping.This rationale also applies when patients are diagnosed with difficult-to-biopsy metastases. 83Furthermore, due to the invasiveness of tissue sampling which can be particularly unpleasant for patients, serial longitudinal tissue biopsies are generally avoided.In such cases, repeat blood sampling over time for ctDNA testing can facilitate longitudinal monitoring of molecular changes in MBC. 11,84A study by Liu et al analysed 300 MBCpaired tissues and liquid biopsies from MBC patients using NGS and found significant concordance between tissue and cfDNA. 85They observed that 77.8% of pathogenic tissue variants were detected in cfDNA, and similarly, 75.7% of pathogenic cfDNA variants were identified in tissue when the tests were conducted within a 7-day interval.However, these percentages decreased to 50.3% and 51.8%, respectively, when the time gap between tests exceeded 365 days. 85ese findings underscore the potential of ctDNA as a non-invasive alternative approach, particularly in situations where tissue availability is limited.Moreover, the study highlights the importance of considering the time interval between tests, as the decrease in concordance over time points out the necessity of timely samplings.
In light of these results, a combined testing approach, incorporating both ctDNA analysis and tissue biopsy, may offer the most comprehensive information by detecting variants that may be missed by either method alone.Therefore, performing ctDNA analysis concurrently with tissue sampling and using it as a substitute for tissue sampling during patient follow-up would be optimal.

| Challenges and optimisation of ctDNA testing
Albeit ctDNA testing seems extremely promising, several intrinsic and technical pitfalls may challenge its integration in MBC management. 86The pre-analytical phase of ctDNA analysis is critical and requires optimisation to ensure reliable testing results. 87This includes the selection of specialised blood collection tubes that allow for an extended time before processing of up to 7 days. 88Centrifugation is another crucial pre-analytical step, as it aids in the removal of formed elements within the blood, including non-neoplastic DNA that may interfere with ctDNA analysis; thus, a double centrifuge is recommended. 89Automated extraction methods are generally preferred over manual extraction in order to minimise human error and ensure consistency. 90Multiple technical factors must be taken into account when assessing the potential for a false-negative test result.These include the quantity of ctDNA under analysis and the sensitivity of the assay, which may vary depending on the specific types of variants being examined.Attention should be paid when obtaining negative results as ctDNA represents only a small portion of cfDNA. 91Indeed, the calculation of tumour fraction, which refers to the proportion of tumour DNA in relation to the total cfDNA, is strictly related to the reliability of the test. 92Tumour fraction can be influenced by several factors such as disease stage (localised vs. metastatic) and overall tumour burden. 935][96] Furthermore, it is important to recognise that the ability of a given test to identify a single nucleotide variant may differ from its ability to detect a structural variant (e.g., gene fusions) or copy number alteration. 83Detection of fusions and copy number alterations require ultrasensitive and comprehensive ctDNA assays; still, in patients with low tumour fraction the rate of confidence is quite low. 97,98Therefore, negative results in ctDNA should be confirmed with analyses on the most representative tumour tissue to ensure accuracy. 99Interpretation of ctDNA results requires careful consideration especially regarding non-tumour mutations that may occur in CHIP.These mutations can overlap with common somatic alterations, 100 thus filtering out CHIP-associated genes is important, particularly for genes such as ATM, BRCA1/2, TP53, and PTEN. 101In terms of gene fusion analyses, ctDNA-based assays have proven essential in complementing tissue testing. 102The advantage of traditional RNA-based assays in comparison to DNA-based assays is that in case of fusion events the RNA expression is often higher and thus the assays are more efficient and sensitive. 103,104Therefore, whether ctDNA should be preferred over cfRNA for these liquid-biopsy based analyses has not yet been established.Further studies that will directly compare DNAand RNA-based assays both on tissue and liquid biopsy samples to test their ability in detecting gene fusions are warranted.Addressing these challenges and optimising the pre-analytical and analytical processes will enhance the reliability and utility of ctDNA testing in routine practice, allowing for more accurate interpretation and informed decision-making in patient management.

| CON CLUS IONS
Liquid biopsy has witnessed significant progress in recent years and has become an important tool in breast cancer management, despite the initial limited application in this setting.However, with the emergence of targeted therapies, ctDNA-based analyses have gained major attention in the identification of resistance mechanisms and in monitoring treatment response in MBC.While tumour tissue remains the gold standard for molecular testing, ctDNA-based approaches can complement or even replace tissue testing in certain situations.The selection of tumour tissue genotyping or liquid biopsy, whether in a sequential or concurrent manner, should be tailored to the unique attributes of each patient and their disease.Furthermore, the elevated cost of an integrative approach adds another layer of complexity for the implementation of the most appropriate diagnostic strategy.In the setting of MBC, frontline tumour tissue testing is recommended for selecting targeted therapies (e.g., PIK3CA mutations) at the time of disease progression, unless the available tissue is inadequate or unavailable.On the other hand, frontline ctDNA analysis may be preferred for detecting acquired resistance mutations (e.g., ERBB2 mutations) or monitoring treatment response (e.g., ESR1 mutations).The timing of blood sampling is crucial to maximise the detection of ctDNA and should be considered carefully.Nonetheless, a negative ctDNA test should be confirmed with an analysis of the most representative tumour tissue sample.
The choice of technical approaches should be guided by the specific clinicopathological features of the patient.NGS may be preferred when a high number of biomarkers needs to be tested.On the contrary, PCR-based methods could be adopted for longitudinal disease monitoring with repeated blood sampling over time.As the field of liquid biopsy continues to evolve, it has the potential to revolutionise personalised medicine and deepen our understanding of cancer biology, potentially translating the acquired knowledge at the early stages of the disease.