Breast cancer is the most commonly diagnosed malignancy among women, comprising approximately 25% of all female cancers. Despite substantial progress in diagnostic techniques and therapeutic interventions, it continues to represent the leading cause of cancer-related mortality in the female population (1). In addition to the classical clinical classification based on ER, PR, HER2 status, breast cancer heterogeneity is more accurately captured by the intrinsic molecular subtypes described by Perou et al. This gene-expression–based taxonomy identifies biologically distinct groups with different behaviors and therapeutic vulnerabilities. The luminal A subtype is characterized by strong hormone-receptor signaling and low proliferative activity, and it is generally associated with a more favorable prognosis. Luminal B tumors also express hormone-receptor–related genes but show higher proliferation rates and, in some cases, HER2 activation, resulting in a more aggressive clinical course. The HER2-enriched subtype is defined by pronounced activation of the HER2 pathway and limited hormone-receptor expression, making it particularly responsive to HER2-targeted therapies.

Finally, the basal-like subtype, which largely overlaps with triple-negative breast cancer, displays high proliferation, genomic instability, and poor prognosis. A fifth group, known as normal-like, resembles non-tumoral breast tissue; however, its clinical relevance remains limited and is sometimes considered a technical artifact.

TNBC is more common in younger women, displays a high proliferative index, and is associated with early relapse, poor prognosis, and increased mortality (1, 2). While chemotherapy remains the standard treatment, response rates vary significantly. Approximately one-third of patients achieve a pathological complete response (pCR) with favorable survival outcomes, whereas the majority experience residual disease and an increased risk of recurrence. Metastatic TNBC has particularly poor survival rates, with over 70% of patients succumbing to the disease within five years (2). Chemotherapy remains the primary treatment for TNBC, but many patients do not achieve complete remission, emphasizing the need for targeted therapies. Key biomarkers under investigation include BRCA1/2 mutations, found in 10%-20% of TNBC cases, and PD-L1 expression, detected in 20%-38% of metastatic cases. Elevated tumor-infiltrating lymphocytes (TILs) correlate with better prognosis, improved immunotherapy response, and higher pCR rates. Histopathological examination remains the gold standard for diagnosis, though limitations such as tumor heterogeneity (2, 3). Genomic profiling has uncovered promising therapeutic targets, including DNA damage response pathways, angiogenesis inhibitors, immune checkpoint inhibitors, and anti-androgen therapies, all of which are currently being tested in clinical trials. However, genetic alterations in TNBC can evolve during treatment, potentially rendering driver mutations functionally neutral and contributing to therapeutic resistance. These challenges underscore the urgent need for personalized therapeutic strategies to improve patient outcomes in TNBC. Although combination chemotherapy improves response rates, TNBC remains a major clinical challenge due to the lack of targeted treatments (3). Research continues to explore ways to better characterize and treat this aggressive form of breast cancer. Metastatic TNBC exhibits significant heterogeneity due to various somatic mutations and molecular alterations, necessitating the identification of novel biomarkers to guide treatment and improve clinical management. Tumor evolution and intratumoral heterogeneity contribute to treatment resistance, driven by cellular mutability and selective drug pressures. While serial biopsies and continuous monitoring could provide insights into tumor progression, their invasive nature limits feasibility (3, 4). Advances in liquid biopsy, particularly circulating tumor DNA (ctDNA) analysis, offer a promising alternative for detecting tumor heterogeneity, monitoring disease progression, and assessing treatment response (4) This review examines the clinical relevance of ctDNA and other liquid biopsy components in breast cancer, with a focus on TNBC. It discusses recent advances in noninvasive technologies for prognostic and predictive assessment, highlighting ctDNA’s potential to guide personalized treatment, monitor response, and improve outcomes. The utility of CTCs, exosomes, and RNA-based biomarkers is also addressed, along with future directions for integrating liquid biopsy into routine clinical care.

TNBC represents around 10-15% of all breast cancer diagnoses. While the overall five-year survival rate is 78.6%, it decreases markedly in advanced stages (III and IV). Treatment is challenging due to the absence of HER2, estrogen, and progesterone receptors, which restricts targeted therapy options. The disease is further complicated by its biological heterogeneity, aggressive nature, and frequent late detection. Moreover, the lack of effective biomarkers hinders early intervention and contributes to unfavorable outcomes (5). In addition to the intrinsic subtypes of breast cancer defined by Perou et al., triple-negative breast cancer (TNBC) itself exhibits substantial molecular heterogeneity. A pivotal study by Jiang et al. analyzed a cohort of 465 primary TNBCs and identified four transcriptome-based subtypes with distinct genomic features and potential therapeutic vulnerabilities.

Clinical decisions for TNBC still primarily depend on the assessment of three markers: ER, PR, and HER2. Advances in ‘omics’ technologies (such as genomics, epigenomics, transcriptomics, and proteomics) have provided deeper insights into the heterogeneity of TNBC, uncovering potentially actionable molecular features in some subtypes (5, 6). On average, TNBC tumors carry 1.68 somatic mutations per megabase of coding regions, resulting in approximately 60 somatic mutations/tumor. Most of these mutations are present at low (<5%) or very low (<1%) allele frequencies, which makes their detection challenging in routine clinical testing. Among the genes involved, BRCA1 and BRCA2 play a pivotal role in maintaining genomic stability, as they are essential for the homologous recombination repair of DNA double-strand breaks. Loss-of-function mutations in these genes compromise DNA repair, leading to genomic instability and increased susceptibility to tumor development. Detecting such rare mutations accurately is critical for identifying patients who may benefit from targeted therapies. Mutations in these genes are found in about 10% of triple-negative breast cancer (TNBC) cases (6, 7). In cells deficient in BRCA1/2, the inhibition of PARP—an enzyme involved in repairing single-strand DNA damage—leads to the accumulation of DNA breaks, ultimately causing cell death. This mechanism is the basis for the use of PARP inhibitors, such as Olaparib, which significantly improved progression-free survival in BRCA-mutated TNBC patients in the OlympiAD trial, compared to standard chemotherapy (7).

Furthermore, the PrECOG 0105 neoadjuvant trial demonstrated that TNBC patients with BRCA1/2 mutations had the highest response rate (75%) to a regimen including gemcitabine, carboplatin, and the PARP inhibitor iniparib. This highlights the potential benefit of targeting DNA repair deficiencies in this patient population (8).

While over 75% of BRCA1-mutated and about 50% of BRCA2-mutated breast cancers exhibit a TNBC phenotype, only 10–20% of all TNBC cases are due to inherited (germline) BRCA mutations. An additional 30–40% show sporadic BRCA1/2 inactivation, often caused by promoter hypermethylation—a phenomenon referred to as “BRCAness”, which mimics hereditary BRCA-related cancers (9).

TNBC also displays dysregulation in the PI3K/AKT signaling pathway, with activating mutations in PIK3CA and AKT1 found in 25-30% of advanced cases. This suggests that combining AKT inhibitors with chemotherapy could improve treatment outcomes. Additionally, Some TNBCs overexpress PDL1, leading to genomic instability and immune infiltration. PD-L1 inhibitors, such as atezolizumab and pembrolizumab, have shown promising results, with atezolizumab receiving FDA approval for metastatic TNBC in 2019 (10). Basal-like Immune Activated (BLIA) has the best prognosis, while Basal-like Immune Suppressed (BLIS) has the worst. Despite significant progress in the molecular profiling of triple-negative breast cancer (TNBC), a comprehensive classification system that integrates genomic, transcriptomic, and histological data has not yet been established (1, 2, 10). This lack of a unified framework, combined with the biologically aggressive nature of TNBC and the limited availability of effective targeted therapies, complicates its clinical management. As a result, TNBC remains a challenging cancer to treat due to its aggressive behavior, the scarcity of reliable biomarkers, and the absence of personalized treatment strategies. The development of an integrated classification system and the identification of novel actionable mutations are critical for advancing precision medicine and improving patient outcomes (1–4).

Liquid biopsy represents a promising diagnostic tool to improve the management and prognosis of TNBC. However, most current techniques are not specifically optimized for TNBC or basal-like tumors, raising concerns about their detection accuracy in these subtypes. This review explores state-of-the-art mutation- and DNA methylation-based liquid biopsies and evaluates their potential applications in TNBC and basal-like cancers (10).

Liquid biopsy has become a promising, non-invasive diagnostic tool for monitoring cancer progression in real-time by analyzing biomarkers like ctDNA and CTCs. ctDNA, in particular, is gaining recognition as a powerful biomarker for improving breast cancer diagnosis, treatment selection, and prognosis. Unlike traditional tissue biopsies, liquid biopsy provides a non-invasive, real-time way to track tumor behavior, detect minimal residual disease (MRD), and identify early signs of recurrence. This is especially valuable for aggressive breast cancer subtypes like TNBC, which have high relapse rates and require precise monitoring (11).

In the context of TNBC, ctDNA analysis holds significant promise for detecting MRD during neoadjuvant chemotherapy, with emerging evidence indicating a correlation between ctDNA levels and patient survival outcomes (12). Particularly in early-stage TNBC, post-neoadjuvant ctDNA assessment has demonstrated potential in predicting recurrence, thereby facilitating earlier and more personalized therapeutic interventions. Compared to traditional biomarkers, ctDNA-based liquid biopsy provides superior sensitivity and clinical utility, especially in recurrent and metastatic settings, where the detection of mutations such as ESR1 and PIK3CA can assist in guiding treatment decisions and evaluating drug response and resistance (13).

Despite promising findings, the clinical use of ctDNA in breast cancer, especially in TNBC, remains limited due to the lack of standardized protocols for detecting recurrence based on genetic alterations (11, 14). Most evidence comes from small-scale retrospective studies. While ctDNA and CTCs show prognostic value, with ctDNA capable of detecting relapse months before clinical signs, routine clinical use is still restricted, particularly for asymptomatic patients (14). Ongoing trials are needed to refine liquid biopsy methods and validate ctDNA’s role in monitoring disease progression and guiding personalized treatments, with a focus on TNBC. Current clinical investigations aim to establish the role of reliable blood-based biomarkers to enhance disease monitoring and improve therapeutic outcomes in breast cancer care (13–15).

While traditional tissue biopsies remain a cornerstone for cancer diagnosis, they are invasive and provide only a snapshot of the tumor at a single point in time, often failing to capture the full spectrum of genetic heterogeneity. In contrast, cfDNA—particularly the fraction originating from tumor cells, known as ctDNA—offers a non-invasive and dynamic alternative for monitoring disease progression (21). Due to its short half-life, ctDNA reflects the current tumor burden, enabling the early detection of genetic alterations, monitoring of treatment response, assessment of minimal residual disease, and informed therapeutic decision-making in breast cancer. Despite its promise as a biomarker, several technical limitations remain. The reliability of ctDNA analysis is highly dependent on proper sample collection and processing, as mishandling can lead to contamination or reduced sensitivity. For optimal results, plasma is preferred over serum, with rapid processing, appropriate storage conditions, and the use of stabilizing collection tubes being strongly recommended (22) Table 1.

Detection methods are broadly categorized into PCR-based and next-generation sequencing (NGS)-based techniques, each with specific advantages and limitations. Emerging ctDNA assays are classified into two categories: tumor-informed and tumor-agnostic assays (23).

PCR-based techniques—such as ARMS-PCR, ddPCR, and BEAMing—offer extremely high sensitivity, capable of detecting mutant alleles at frequencies as low as 0.001% (1 in 100,000 wild-type molecules). However, their major limitation lies in requiring prior knowledge of the target mutation, which restricts their utility in tumors with poorly characterized genetic profiles. In breast cancer studies, particularly in metastatic settings, PCR-based detection of mutations in PIK3CA and TP53 has shown good concordance with tissue-based findings and has been correlated with treatment response and disease progression. Notably, ddPCR has also enabled the detection of tumor-specific chromosomal rearrangements, with detection limits around 0.01% (25).

In contrast, NGS-based approaches allow for broader mutational screening, particularly useful in identifying non-recurrent or novel mutations such as those in TP53 or PTEN (25, 26). Techniques like Techniques like tagged-amplicon deep sequencing (TAm-Seq) and Safe-Sequencing System (Safe-SeqS) improve the analytical sensitivity of NGS by minimizing sequencing errors and enabling reliable detection of ctDNA even at low allelic fractions (<1%). Recent studies in TNBC patients have demonstrated a high concordance between tumor and plasma DNA for TP53 mutations, with mutant allele fractions ranging from 2% to 70% (median 5%). However, NGS sensitivity remains insufficient for clinical application in early-stage or minimal residual disease, where ctDNA concentration is typically low (26).

Importantly, in TNBC patients with residual disease after neoadjuvant chemotherapy (NAC), ctDNA detection via NGS has been associated with increased relapse risk. In a cohort of early-stage TNBC patients, post-treatment ctDNA positivity was predictive of disease recurrence, highlighting the potential of ctDNA as a biomarker for risk stratification and treatment tailoring in the post-NAC setting (25, 26) (Figure 1).

Molecular studies are opening new avenues for the clinical management of TNBC, with circulating cell-free tumor nucleic acids (ctNAs) representing one of the most promising research areas.

A major advancement in the clinical application of ctNAs is the possibility of isolating them through liquid biopsy, a non-invasive procedure that analyzes various biological fluids — including blood, urine, cerebrospinal fluid, and pleural effusion — for tumor-derived components (27).

Liquid biopsy allows the detection and monitoring of cancer in real time by isolating CTCs, circulating cell-free nucleic acids (ccfNAs), exosomes, and tumor-educated platelets (TEPs). These components reflect the genetic and phenotypic heterogeneity of both primary tumors and metastatic lesions, making them valuable as predictive and prognostic biomarkers (28).

Despite their potential, ctNAs are typically present in low concentrations and are often highly fragmented, which poses technical challenges for clinical use. Therefore, the development of sensitive and specific analytical platforms, such as digital droplet PCR and NGS, as well as optimized workflows for ctNA isolation and preservation, are essential for their reliable use in routine clinical practice (Figure 2).

Non-coding RNAs (ncRNAs) are increasingly recognized as key regulators of gene expression and cancer progression (28, 29). Despite lacking protein-coding capacity, ncRNAs — including microRNAs (miRNAs), long non-coding RNAs (lncRNAs), circular RNAs (circRNAs), and PIWI-interacting RNAs (piRNAs) — modulate transcriptional, post-transcriptional, and epigenetic processes. These molecules influence fundamental cellular behaviors such as proliferation, differentiation, apoptosis, and immune escape. Aberrant ncRNA expression has been implicated in tumorigenesis across various cancer types, including breast cancer (BC), where they are now under active investigation as potential diagnostic and prognostic biomarkers (29).

Among ncRNAs, microRNAs (miRNAs) have drawn particular attention for their potential role in liquid biopsy applications. Tumor-derived miRNAs can be released into the bloodstream, reflecting molecular alterations in cancer cells and offering a non-invasive window into tumor dynamics (29, 30). Their remarkable stability in circulation, coupled with the ability to capture information from various body fluids, positions miRNAs as promising biomarkers for cancer detection and monitoring.

Extracellular vesicles (EVs), including exosomes and microvesicles, are membrane-bound structures secreted by cells that encapsulate a broad spectrum of bioactive molecules such as DNA, miRNAs, mRNAs, proteins, and lipids. These vesicles are pivotal mediators of intercellular communication and contribute to key oncogenic processes including metastasis, angiogenesis, immune evasion, and therapy resistance. In TNBC, tumor-derived exosomal miRNAs—particularly miR-21, miR-1246, and miR-939—have been implicated in tumor aggressiveness, lymph node involvement, and disease-free survival, highlighting their potential as prognostic biomarkers (35). Recent studies and meta-analyses underscore the diagnostic and predictive utility of exosome-associated molecules in breast cancer. Circulating exosomes, owing to their remarkable stability in bodily fluids, represent a valuable substrate for liquid biopsy approaches. Exosomal cargo, including tumor-associated proteins (e.g., CEA, CA15-3) and distinct miRNA signatures, mirrors the molecular landscape of the primary tumor and has shown correlations with tumor subtype, response to neoadjuvant chemotherapy, and risk of recurrence in TNBC. While ctDNA remains the gold standard for genomic mutation profiling due to its higher sensitivity and specificity, exosomal biomarkers provide complementary insights and are actively investigated for their clinical applicability in TNBC diagnostics, prognostication, and therapeutic monitoring (36).

In the early stages of breast cancer (BC), when the malignancy remains confined to the breast and regional lymph nodes, ctDNA levels in the bloodstream are generally low, reflecting a limited tumor burden. This poses a challenge for ctDNA detection, particularly in TNBC, where the genomic landscape is frequently undefined. For this, broad genomic strategies — including targeted sequencing panels and whole-exome sequencing (WES) — are recommended in these cases, provided that the assay sensitivity is adequate to detect low-abundance ctDNA (37).

Despite the anatomical accessibility of breast tumors for percutaneous biopsy, recent evidence has highlighted the clinical relevance of ctDNA analysis in early-stage disease. For example, PIK3CA mutations, commonly observed in breast cancer, were successfully detected in plasma samples with high specificity and sensitivity. A study involving 29 patients reported a 93% concordance between PIK3CA mutations detected in tumor tissues and matched pre-surgical plasma samples, underscoring the technical feasibility of ctDNA detection using digital droplet PCR (ddPCR) in early disease (38).

In addition, TP53 mutations are frequently associated with BRCA1-mutated breast cancers. The ongoing prospective CirCA01 trial aims to evaluate the potential of ctDNA as an early diagnostic and surveillance biomarker in asymptomatic BRCA1 mutation carriers. Specifically, the study investigates whether serial monitoring of TP53 mutations in plasma can enable early detection of malignancies, predict relapse, and distinguish new primary tumors from recurrent disease.

In early TNBC, both ctDNA and circulating tumor cells (CTCs) have demonstrated prognostic and predictive value, particularly for early identification of minimal residual disease (MRD) and relapse risk following neoadjuvant chemotherapy (NAC). The Q-CROC-03 trial showed that the persistence of ctDNA after NAC was independently associated with an increased likelihood of relapse, while ctDNA clearance was correlated with favorable long-term outcomes comparable to patients who achieved pCR (39).

Similarly, the I-SPY 2 trial demonstrated that ctDNA persistence following neoadjuvant chemotherapy is significantly correlated with disease recurrence, while ctDNA clearance predicts favorable prognosis (40).

These findings support the potential role of ctDNA as a non-invasive tool for early risk stratification and for tailoring post-treatment management strategies. These observations highlight the potential of ctDNA-based assays to refine therapeutic decision-making and personalize patient monitoring. However, technical limitations, particularly in early-stage disease, and the lack of standardized clinical implementation still pose significant challenges for routine use.

Metastatic TNBC is typically diagnosed after the emergence of new clinical symptoms or through imaging when metastases have become sufficiently advanced. Detecting metastasis at an earlier stage could significantly enhance the prognosis for patients with stage IV TNBC(). Olsson et al. highlighted the potential of ctDNA analysis for the earlier detection of metastatic disease in patients who had previously undergone treatment for early-stage BC (48). As previously mentioned, chromosomal rearrangements were successfully identified by ddPCR in plasma samples, on average, 11 months before the first clinical signs of metastasis appeared. Additionally, ctDNA could serve as a reliable tool to confirm metastatic relapse when imaging fails to definitively detect metastases. This approach would negate the need for a solid biopsy, as the mutation responsible for metastasis development would have been identified in prior tumor analysis. However, further studies are required to validate the full potential of ctDNA in such diagnostic application (48, 49) (Figure 3).

Identifying abnormal DNA methylation signatures through non-invasive methods offers a promising avenue for the early detection of TNBC, paving the way for earlier clinical intervention and more tailored therapeutic approaches (50).

In several studies, genome-wide DNA methylation data were analyzed from TNBC tumor tissues, normal breast tissues, peripheral blood samples of TNBC patients and healthy controls, as well as reference samples of sorted blood and mammary cells.

Through stringent selection, verification, and external validation, the researchers identified 23 differentially methylated regions (DMRs) associated with TNBC. A machine learning algorithm was then used to prioritize the most promising DMRs, and six were selected for further evaluation in plasma-derived cell-free DNA (cfDNA) samples (50, 51).

Using droplet digital PCR, the presence of methylation at these regions was confirmed in cfDNA samples from TNBC patients compared to healthy controls. A combined methylation score based on three specific DMRs (associated with the SPAG6, LINC10606, and TBCD/ZNF750 genes) showed the strongest potential for discriminating TNBC patients from controls, achieving an AUC of 0.78 in the test set and 0.74 in the independent validation set (52).

These findings suggest that cfDNA methylation signatures may serve as sensitive and specific non-invasive biomarkers for the early detection of TNBC, highlighting the clinical potential of liquid biopsy for cancer screening, diagnosis, prognosis, and therapeutic monitoring.

However, the authors emphasize that further large-scale and independent studies are warranted to validate these biomarkers before clinical implementation.

Identifying reliable tumor-derived predictive biomarkers for triple-negative breast cancer (TNBC) patients receiving immunotherapy remains a significant challenge. Emerging approaches—such as monitoring T-cell clonal expansion and analyzing ctDNA—are showing potential. Meanwhile, the prognostic relevance of CTC quantification, especially in relation to PD-L1 expression on CTCs, is currently under active investigation (57).

Tumor Mutational Burden (TMB), which reflects the number of somatic mutations, could serve as a marker for neoantigen production and T-cell infiltration. In breast cancer (BC), TMB’s role in tumor immunogenicity is uncertain, as it is typically low to intermediate, with TNBC showing the highest median TMB (57, 58). While preliminary data suggests that hypermutated BC may benefit from PD-1 inhibitors, TMB-H is not consistently predictive for immune checkpoint inhibitors (ICIs) across cancer types. Only a small percentage of TNBC cases (8-10%) exhibit high TMB. Researchers propose refining TMB’s predictive value in TNBC by considering specific mutational signatures and other genomic alterations.

Liquid biopsy-based TMB is under investigation, and ctDNA analysis could serve as a clinically relevant biomarker. In various trials, ctDNA molecular response has been linked to survival, with ctDNA dynamics showing stronger correlations. The NIMBUS trial showed that HER2-negative mBC patients with TMB-H benefited from nivolumab plus ipilimumab, with certain patients showing higher response rates, emphasizing the need for optimized TMB cut-off values (58). The SAFIR02 BREAST IMMUNO study identified CD274 (PD-L1 gene) as a new biomarker linked to better responses from ICIs in metastatic TNBC, suggesting new opportunities and challenges for liquid biopsy applications (59). The tumor microenvironment in breast cancer (BC) is characterized by a complex interplay of immunoregulatory molecules and infiltrating immune cells, offering insight into the host’s antitumor immune activity. In triple-negative breast cancer (TNBC), liquid biopsy technologies are proving instrumental in uncovering spatial heterogeneity of PD-L1 expression between primary and metastatic lesions.

Elevated PD-L1 levels on CTCs have been correlated with shorter overall survival in patients with metastatic breast cancer (mBC). Additionally, the presence of immunosuppressive factors such as early myeloid-derived suppressor cells (eMDSCs) and their associated proteins has been implicated in suboptimal responses to neoadjuvant chemotherapy in TNBC.

Additionally, prior studies across all BC subtypes reported elevated levels of CD117+ and CD11b+ granulocytes at diagnosis, which decreased after tumor removal by surgery and radiotherapy — supporting the link between tumor burden and measurable immune response (58–60).

A reduction in serum concentrations of MDSCs, PD-L1-expressing T lymphocytes, and regulatory T cells (Tregs) has been correlated with clinical benefit in BC.

Moreover, tumor-associated macrophages (TAMs), particularly CD163-positive subsets, play a crucial role in tumor progression and immune evasion. Monitoring circulating TAM levels could represent a future serum-based immune biomarker to assess treatment response and potentially guide macrophage-targeted therapies in TNBC patients (60).

TNBC is an aggressive subtype with poor prognosis and limited treatment options. Often diagnosed at an advanced stage, it highlights the urgent need for reliable, non-invasive biomarkers to enable early detection, guide therapy, and improve patient outcomes (1, 61).

This narrative review confirms the prognostic value of ctDNA in early TNBC, particularly around neoadjuvant therapy (NAT). ctDNA detection—especially after NAT or surgery—is consistently associated with worse disease-free and overall survival, identifying patients at high risk of relapse (40, 62). Studies by Garcia-Murillas, and Nader-Marta emphasize the value of ctDNA for early risk stratification and personalized treatment decisions. While pCR remains important, early ctDNA clearance during NAT—regardless of pCR—is linked to better outcomes (62).

Compared to CTCs, ctDNA is more sensitive and easier to detect, making it a more reliable marker of minimal residual disease (MRD). Nonetheless, limitations remain. Some studies, such as those by Chen et al. (2024), Cavallone et al. (2020), and the TBCRC 030 trial, report inconsistent predictive accuracy, including false positives and undetectable ctDNA in relapsing patients. Technical variability, low ctDNA shedding in early disease, and lack of assay standardization also hinder its reliability (21, 63).

These conflicting findings are largely due to technical limitations and differences in patient populations. In early TNBC, ctDNA levels are often very low, making detection difficult. For example, Parsons et al. reported a median ctDNA allele fraction of just ~3.7×10^–4 post-NACT, with 58% of patients having levels below assay sensitivity. Thus, “undetectable” often reflects test limitations rather than true absence of disease. Conversely, detectable ctDNA may originate from small residual clones that can still be eliminated by adjuvant therapies. Despite these challenges, ctDNA provides additional prognostic information beyond traditional markers like pCR. In particular, ctDNA status before and after surgery correlates with pCR and residual cancer burden (RCB), refining risk assessment. Ongoing trials (e.g., CUPCAKE and Apollo) are evaluating ctDNA-guided strategies, though their effectiveness will depend on available therapeutic options (64, 65).

Treatment context also matters. Many early TNBC patients now receive post-NACT therapies like capecitabine or immunotherapy, which improve outcomes even in ctDNA-positive cases. In fact, pooled data show that non-pCR patients now reach 62–70% five-year disease-free survival with modern adjuvant treatment—suggesting that ctDNA positivity no longer reliably predicts relapse on its own. The I-SPY2 trial and Cailleux et al. further support early ctDNA clearance (within three weeks of NAT) as a predictor of pCR and reduced RCB (66). This suggests a role for dynamic ctDNA monitoring in real-time treatment adaptation. Future improvements—such as using multiple gene targets in ddPCR—could enhance detection sensitivity. Specific ctDNA alterations—such as 17q22 amplification—have been linked to enhanced cisplatin sensitivity due to impaired DNA repair. Other mutations (e.g., TP53, PIK3CA/AKT pathway, NF1, PTEN, CHEK2) are associated with progression or resistance and may offer future therapeutic targets, although most are still under clinical investigation.

Although promising, ctDNA is not yet fully validated as a clinical tool in early TNBC. Prospective, multicenter trials (e.g., BRE12-158) are needed to confirm its prognostic and predictive utility across subtypes. Notably, ctDNA outperforms CTCs in early TNBC, though CTCs may retain relevance in other breast cancer types (67). DNA methylation also plays a critical role in the regulation of gene expression in TNBC, contributing to tumor progression and metastasis. In a recent study, methylation alterations were identified in 16 out of 38 TNBC-specific genes and were linked to changes in expression and lymph node involvement. These findings suggest that epigenetic profiling may uncover novel biomarkers for prognosis and guide the development of targeted strategies in this aggressive breast cancer subtype (68).

Overall, the integration of serial ctDNA monitoring into the neoadjuvant management of TNBC holds considerable promise for enhancing risk stratification and enabling timely, personalized interventions. Persistent or re-emerging ctDNA after therapy—regardless of pCR—has been consistently associated with an increased risk of early relapse, underscoring its prognostic value beyond conventional markers. As such, ctDNA represents a powerful tool to refine post-treatment surveillance and guide therapeutic decision-making. Future clinical trials should focus on evaluating ctDNA-informed treatment escalation strategies, particularly for molecular non-responders, in order to improve long-term disease control and survival outcomes in this high-risk population (69–71).

Liquid biopsy represents a promising, non-invasive tool for monitoring treatment response and disease progression in TNBC. Techniques such as ctDNA and CTCs have demonstrated potential for early detection, prognosis, and therapy guidance, especially in identifying residual disease and predicting recurrence. Notably, ctDNA mutations, including TP53 and BRCA1, may serve as prognostic biomarkers in early-stage TNBC. However, the clinical utility of these biomarkers remains limited by technological variability, lack of standardization, and insufficient large-scale validation. Future efforts must focus on refining detection methods, conducting robust clinical trials, and integrating multi-parametric liquid biopsy approaches—including RNA, exosomes, and immune biomarkers—into personalized treatment strategies to improve patient outcomes in TNBC.

While liquid biopsy represents a transformative approach for the management of triple-negative breast cancer (TNBC), several limitations currently hinder its widespread clinical integration. First, the sensitivity of ctDNA and CTC detection can be compromised in patients with minimal residual disease or low tumor burden, potentially resulting in false-negative results. Conversely, false positives may arise from clonal hematopoiesis of indeterminate potential (CHIP) or other non-tumor-derived alterations, necessitating careful interpretation of results.

Heterogeneity in assay platforms, pre-analytical handling, and bioinformatic pipelines further complicates cross-study comparisons and limits reproducibility. In TNBC specifically, high intra- and inter-patient heterogeneity poses an additional challenge, as a single liquid biopsy sample may not fully capture the evolving tumor genomic landscape. Furthermore, most of the available data derive from small, retrospective, or early-phase studies, which may be subject to selection bias and lack long-term follow-up.

To address these constraints, future research should prioritize prospective, large-scale, and longitudinal studies with standardized methodologies, integrating multi-omic liquid biopsy analyses and correlating them with robust clinical endpoints. The incorporation of artificial intelligence–driven algorithms to interpret complex, multi-analyte datasets could enhance predictive accuracy and facilitate real-time therapeutic decision-making. Additionally, cost-effectiveness analyses will be critical to ensure equitable access and adoption in routine oncology practice.

Ultimately, overcoming these limitations will be essential to fully realize the potential of liquid biopsy as a reliable diagnostic, prognostic, and treatment-monitoring tool in TNBC.