Liquid Biopsy for Colorectal Cancer: Advancing Detection and Clinical Application

Abstract

Colorectal cancer (CRC) is the third most common cancer and a leading cause of cancer-related mortality worldwide, with prognosis significantly deteriorating at advanced stages. While current diagnostic methods, such as colonoscopy and tissue biopsy, are widely employed in clinical practice, they are invasive, expensive, and limited in assessing tumor heterogeneity and monitoring disease processes, including therapy response. Therefore, early and accurate detection, coupled with minimal invasion and cost-effective strategies, are critical for improving patient outcomes. Liquid biopsy has emerged as a promising, minimally invasive alternative, enabling the detection of tumor-derived components. This approach is increasingly utilized in clinical settings. The current key liquid biopsy modalities in CRC include circulating tumor DNA (ctDNA), circulating tumor cells (CTCs), and RNA-based biomarkers such as long non-coding RNAs (lncRNAs), microRNAs(miRNAs), and circular RNAs (circRNAs), and tumor-educated platelets (TEPs). These methods provide valuable insights into genetic and epigenetic tumor alterations, and serve as indicators for early detection, treatment monitoring, and recurrence prediction. However, challenges such as assay standardization and variability in sensitivity persist. This review delves into the clinical applications of liquid biopsy in CRC management, highlighting the transformative roles of ctDNA, CTCs, and non-coding RNAs, TEPs in early detection, prognostic assessment, and personalized therapy. In addition, it addresses current limitations and explores potential advancements to facilitate their integration into routine clinical practice.

1. Introduction

Colorectal cancer (CRC) is the third most common cancer and second leading cause of cancer deaths worldwide [1]. In 2022, there are more than 1.9 million new cases with 904,000 deaths worldwide [1]. The 5-year relative survival rate of metastatic colorectal cancer is 15% [2], which is much lower than the rates of 90% in localized diseases and 70% in regional diseases [2]. Therefore, early detection and intervention hugely improve the patient survival rate [3]. The current screening and diagnosis of CRC include colonoscopy, flexible sigmoidoscopy, fecal occult blood test, fecal immunochemical test, multitargeted stool DNA test (Cologuard), CT colonography [4,5,6], and treatments include endoscopic and surgical excision for primary and metastatic disease, chemotherapy, immunotherapy, and targeted therapy [4,5].

Non-invasive serum tumor biomarkers such as carcinoembryonic antigen (CEA) and carbohydrate antigen 19-9 (CA19-9), as well as multitarget biomarkers in stool including tumor-associated DNA, mRNA, protein, and intestinal microbial flora changes, are also commonly used in the diagnosis of colorectal cancer, but those markers have low sensitivity and specificities, often necessitating confirmation via colonoscopy biopsy, which is still the golden standard method for diagnosis of CRC [6,7,8,9,10,11,12,13,14]. However, colonoscopy biopsy is invasive and requires extensive patient preparation, and it carries risks such as bowel tears and bleeding. Furthermore, it is associated with high costs and several technical complexities, including limited capability to assess spatial tumor genetic heterogeneity, insufficient specimens, difficulty in longitudinal monitoring, as well as the need for well-trained personnel to perform the tests [6,8,15,16].

Since CRC is preventable and curable when diagnosed early, the early detection and intervention of CRC will make a huge difference in mortality for the patients [2,4]. It is crucial to identify non-invasive tools with high specificity and sensitivity to aid in early detection, prognosis, and treatment monitoring [6].

In recent years, one minimal invasive method, also known as liquid biopsy including circulating tumor cells (CTCs), circulating tumor DNA (ctDNA), microbiome, exosomes, circulating non-coding RNAs (microRNA, long non-coding RNAs, etc.), tumor-educated platelets (TEPs), etc. [6,17,18,19], has gained significant prominence [17,18,20,21].

This review explores the utility of liquid biopsy in CRC patients, focusing on key biomarkers such as CTCs, ctDNA, circulating non-coding RNAs, and tumor-educated platelets. It emphasizes their clinical relevance in early detection, monitoring minimal residual disease (MRD), and guiding treatment strategies. Additionally, the review addresses current limitations and discusses potential advancement to enhance the future implementation in CRC management.

2. Circulating Tumor Cells

Circulating tumor cells (CTCs) are tumor cells that detach from the primary tumor, enter the bloodstream, and travel to the metastatic sites [22,23]. During the process, epithelial–mesenchymal transition (EMT)-related CTCs play a crucial role, as epithelial CTCs transition into mixed CTCs and eventually acquire a mesenchymal phenotype, enhancing their invasive and metastatic potential [24,25]. In addition to the mesenchymal-correlated CTCs and mesenchymal-types of CTCs, circulating cancer stem cells (CCSCs), known for their self-renewal capability, and CTC clusters, which form when cancer cells aggregate either alone or with stromal or immune cells, also play significant roles in initiating metastases [25]. With a median half-life of 1–2.5 h in circulation, CTCs represent a heterogeneous population of cancer cells with diverse phenotypes and functions, collectively driving the metastatic process, and their presentation is closely associated with a poor prognosis [25].

CTCs are particularly prevalent in advanced stages of CRC, making them valuable for monitoring cancer progression, CRC recurrence, and guiding therapy strategies [22,26,27]. He et al. reported that CTCs, like CEA, CA125, and CA199, serve as an independent risk factor for CRC metastasis prediction [26]. CTCs may serve as a more reliable indicator of CRC recurrence compared to CEA, as Wang reported that positive CTC results are identified in patients following radical surgery, while their perioperative serum CEA level are normal [25,28]. Other studies also demonstrated positive CTCs as having higher sensitivity and specificity in predicting postoperative recurrence than CEA, approximately 8 months earlier than elevated CEA (median 10.7 vs. 2.8 months), and its positivity is strongly associated with poor overall survival (OS) [25]. Additionally, the persistent CTC positivity, defined as CTC detection at two or more times, is a prognostic factor for early tumor recurrences in stage III CRC patients, both in postoperative (within one year following radical resection) CRC patients and after chemotherapy (adjuvant mFOLFOX) [25,29,30].

3. Methods for Detecting CTCs

The CTC blood concentration is extremely low, with metastatic patients typically having only 1–10 CTCs (either single or in clusters) per milliliter of blood, presenting significant challenges for the isolation process [31]. Furthermore, the heterogeneity of CTCs adds another layer of complexity to detecting all the CTC subpopulations in circulation. Conventional CTC identification methods have traditionally relied on markers like the epithelial cell adhesion molecule (EpCAM) or specific mRNA expression, focusing only on conventionally defined CTCs. This approach overlooks other subpopulations, such as epithelial–mesenchymal transition (EMT)-related CTCs, which lose the EpCAM expression and evade detection [25,32]. With advancements in technology, the detection and enumeration of CTCs in circulation have significantly improved, using more biomarkers [25,32]. For example, CellSearch (San Diego, CA, USA) is the first and only Food and Drug Administrated (FDA)-approved system for the isolating, identifying, and counting of CTCs utilizing multimaker staining (EpCAM+, CK+, CD45−, and DAPI+), and it is a highly reproducible method that correlates with the CRC stage [25,32].

An alternative approach for isolating CTCs leverages the size difference between cancer cells and non-malignant blood cells. This method, known as isolation by size of epithelial tumor cells (ISET), focuses on the biophysical properties of cancer cells to enrich the sample. Compared to the CellSearch system, the ISET filtration technique can handle larger sample volumes and isolate a greater number of CTCs, enabling more extensive functional or genomic analyses [14,33]. Additionally, highly sensitive capture methods, such as those utilizing VAR2CSA protein to bind cancer cells to the magnetic bead, has been developed [14,34]. CTCs SE-iFISH further eliminates the reliance on epithelial markers by employing subtractive enrichment, followed by immunostaining and fluorescence in situ hybridization (FISH), for the identification of CTCs [14,18,35]. In addition, CTC-whole exome sequencing (WES) and whole genome sequencing (WGS) provide valuable insights by uncovering the extensive the heterogeneity of cancer [18].

4. ctDNA in CRC

Circulating cell-free DNA (cfDNA) in blood plasma, comprising extracellular nucleic acid fragments released during processes such as cell apoptosis, necrosis, or other cellular activities, has demonstrated significant promise for early cancer detection and as a guideline for cancer treatment and early detection of recurrence [17,18,36].

In cancer patients, a small portion (<1% [37,38,39,40]) of cfDNA is originally from primary tumors, known as circulating tumor DNA (ctDNA), which carry the same mutations and genetic alterations as the primary tumor [17]. These genetic and epigenetic changes in ctDNA, including but not limited to k-RAS mutation [41], gene amplifications [42,43], loss of heterozygosity [44,45], promoter hypermethylation [46,47,48], single-nucleotide mutations [39,49,50,51,52], and cancer-derived viral sequences [53,54,55,56] could be detected by improving assay techniques with high sensitivity [37]. The levels of ctDNA is associated with tumor burden and patient prognosis [17], and it is detected earlier in advanced stage cancers, especially metastatic disease. In CRC patients, ctDNA can be detected in 50% of patients with non-metastatic disease and 90% of patients with metastatic disease [37,57]. Also, the ctDNA detection in CRC patients before and after treatment is associated with higher relapse rate after surgery and adjuvant chemotherapy treatment [18,58,59]. Tie et al. reported in a multicenter cohort study of 96 patients with stage II colon cancer in Australia that ctDNA was detected in 20 patients (17%) in the post-adjuvant chemotherapy samples, and the estimated 3-year recurrence-free interval in patients with detectable ctDNA was 30%, compared to the 77% of patients with undetectable ctDNA (HR, 6.8; 95% CI, 11.0–157.0; p < 0.001) [18,59].

With the improved techniques, the analysis of ctDNA in the plasma provides a minimally invasive method to diagnose, characterize, and monitor the disease the CRC patients [18,60]. ctDNA analysis is quantitative, for it can reflect the actual tumor burden and the change in ctDNA levels during chemotherapy and/or immune checkpoint inhibitor (ICI) therapy, thus potentially serving as the prognostic and predictive biomarker [18,60]. Diehl et al. reported that mutated genes including APC, KRAS, TP53, and PIK3CA were found in ctDNA samples in patients with CRC, and those ctDNA levels showed a significant decrease after surgery and immediate rise after chemotherapy discontinued [18,40].

ctDNA can be used to detect minimal residual disease (MRD), which is defined as a small number of cancer cells persisting in the body after treatment that cannot be detected with current medical imaging modalities [61]. MRD can only be detected using highly sensitive laboratory methods that can find one single cancer cell among one million normal cells like ctDNA and flow cytometry, the latter of which is mostly used in immunology and hematology.

Another clinical application of ctDNA is used to monitor the efficiency of tumor treatment and tumor dynamics [60] such as early detection of recurrence [62] and some resistance mechanisms in KRAS and BRAF mutation [62]. Cao et al. also reported that after treatment with bevacizumab and cetuximab combined with chemotherapy in advanced CRC patients, the mutations such as BRAF, KRAS, AMER1, and other major driving genes showed a reduction in burden compared with the baseline [63]. Additionally, patients with KRAS and TP53 mutations benefit more than their wild-type counterpart [63].

Standardized ctDNA assessment in interventional clinical trials across entities demonstrate the clinical utility of ctDNA as a biomarker for personalized cancer immunotherapy [60]. ctDNA has first been approved for use in clinical practice to detect EGFR mutation in carcinoma in 2014 [18]. Since then, numerous trials have been conducted to explore its applications further. According to the ClinicalTrials.gov, there are currently 242 clinical trials using ctDNA to predict and monitor various tumor (breast, endometrial cancer, lung cancer, gastric cancer, etc.) dynamics, among which, 42 trials are applied in the colorectal cancer [64].

For example, one observational study, GALAXY, found out that MSD detected by ctDNA at 4 weeks post-surgery is the strongest prognostic risk factor for recurrence in patients with resected colorectal cancer, and it is independent of BRAF V600E status and microsatellite instability-high (MSI-H) status [65]. Therefore, the ctDNA analysis post-surgical may guide the treatment plan by stratification of post-surgical risks [66].

With numerous clinical trials underway, methylation-based biomarkers have emerged as a promising tool for the non-invasive detection of colorectal cancer (CRC) [67,68] Among these, methylated SEPT9 (mSEPT9) has been one of the most extensively studied and clinically validated biomarkers. Notably, the Epi proColon^® test, which detects methylated SEPT9 DNA in circulating tumor DNA (ctDNA) isolated from plasma, has been approved by the U.S. Food and Drug Administration (FDA) as a blood-based screening test for CRC [6,67,68]. In addition to SEPT9, other methylation-based biomarkers have also been approved or are under advanced clinical evaluation for colorectal cancer (CRC) detection. For example, a combination of NDRG4 and BMP3 is used in the FDA-approved Cologuard^® test, which integrates methylation markers along with other molecular features for non-invasive CRC screening [69]. Furthermore, biomarkers such as SDC2, VIM, APC, MGMT, SFRP1, SFRP2, and NDRG4 have been identified as key methylation targets, each playing distinct roles in pathways linked to tumor initiation and progression, including Wnt signaling, cell cycle regulation, and DNA repair [69]. These methylation biomarkers reflect epigenetic alterations that occur early in CRC development, making them valuable tools not only for early detection but also for risk stratification and monitoring treatment response.

5. Methods for Detecting ctDNA

The methods of ctDNA could be categorized into either PCR-based or Next-Generation Sequencing (NGS) based techniques [18,37,70]. PCR-based techniques, including digital PCR (dPCR) [71], allele-specific amplification refractory mutation system (ARMS) PCR [72], allele-specific PCR (AS-PCR) [73], droplet digital PRC (ddPCR), or beads, emulsion, amplification, magnetics (BEAMing) [40], can be applied to detect a single or small number of specific known mutations with a high sensitivity [17,18,37,70]. NGS-based methods, including tagged-amplicon deep sequencing (Tam-Seq) [52], Safe-sequencing system (Safe-SeqS) [74], and personalized profiling by deep sequencing [75], could be used to sequence the entire genome and multiple rare mutations simultaneously without information from primary tumor sequencing [17,18,37,70]. The NGS can be utilized in untargeted techniques, such as whole genome sequencing (WGS) and whole exome sequencing (WES), to identify novel, clinically significant genomic aberration without the need for information about the primary tumor [17,18,37,70]. When comparing PCR-based and NGS-based techniques, PCR-based methods are cost-effective and fast, but they can only identify a limited number of prespecified mutations. In contrast, NGS-based techniques can detect a broader range of mutations and analyze multiple genomic targets and alterations. However, NGS is limited by lower sensitivity, higher sample volume requirement, and more expensive and time-consuming procedures [18,76,77].

6. Circulating Non-Coding RNA in CRC

Non-coding RNA (ncRNAs), including but not limited to small interfering RNAs (siRNAs), small nucleolar RNAs (snoRNAs), microRNAs (miRNAs), circular RNAs (circRNAs), and long non-coding RNAs (IncRNAs), etc., constitute over 90% of human genomes but not encoded protein [78,79,80,81,82]. Studies have demonstrated ncRNAs are stable in the serum and their aberrant expressions play an important role in the CRC development like proliferation, differentiation, migration, angiogenesis, and apoptosis [78,79]. ncRNAs, especially, lncRNAs, miRNAs, and circRNAs, are particularly noteworthy for their critical role in monitoring the CRC progression [83].

7. LncRNAs in CRC

The importance of lncRNAs in CRC was first highlighted in a report that the loss of imprinting of long QT intronic transcript 1 (LIT1/KCNQ1OT1) was frequently associated with CRC patients, indicating a potential connection between lncRNAs and CRC [84,85]. Since then, numerous studies have investigated the dysregulated expression of lncRNAs during colorectal cancer development. A growing body of evidence now highlights the biological and clinical significance of specific lncRNAs in CRC. Table 1 provides a summary of potential lncRNAs utilized in CRC screening, diagnosis, prognosis, and treatment.

For example, lncRNA TINCR and lncRNA PVT1 have been reported to exhibit significantly higher levels compared to CEA in early CRC detection, highlighting its potential role as biomarkers for diagnosis of early CRC [86,87,88]. Furthermore, lncRNA may correlate with tumor burden in CRC, as Xu et al. demonstrated that the expression levels of lncRNA ZFAS1, SNHG11, LINC00909, and LINC00654 are significantly reduced in postoperative CRC samples compared to preoperative samples [89]. Additionally, Xu et al. reported that upregulation of DSCAM-AS1 in CRC is positively associated with advanced stages of CRC and metastasis status, suggesting it promotes CRC cell proliferation and migration by targeting the miR-137/Notch-1 axis [90]. Similarly, lncRNA RP11-296E3.2 is highly expressed in metastatic CRC and is linked to short overall survival, and demonstrates superior sensitivity and specificity compared to plasma CEA as a biomarker [87,91]. High expression of LINC00858, MF12-AS1, BANCR, XIST, HOTAIR, etc., are closely associated with advanced stages and poor prognosis of CRC [86,87,88,92,93,94].

Table 1. Summary of potential lncRNA implicated in colorectal cancer, including dysregulation, potential function and mechanism, signaling pathways.

LncRNA	Dysregulation	Potential Function and Mechanism	Signaling Pathways	References
CCAL	Up	Suppress activator protein 2α and activate Wnt/ β-catenin pathway	Wnt/β-catenin	[95]
CCAT 1	Up	N/A	CCAT1/MYC	[96]
CCAT 2	Up	Stabilize and induce expression of BOP1 an activator of aurora kinase B	AURKB	[35]
CAHM	Down	N/A	N/A	[97]
CRNDE	Up	Suppress apoptosis	Wnt/β-catenin	[98,99,100]
H19	Up	mir675 regulate RB	N/A	[101]
HOTAIR	Up	Promote cell invasion, participates in EMT, metastasis	N/A	[102,103]
Loc285194	Down	Regulated by p53, inhibits tumor cell growth partially via repression of mir-211	N/A	[29,104]
LincRNA-p21	Down	Collaborates with hnRNP-K to act as a transcriptional coactivator of p53	N/A	[105]
LIT1/KCNQ1OT1	Loss of Imprinting	Associate with epigenetic status of KvDMR1	N/A	[84,106]
Lnc-LALC	Down	Inhibit expression of LZTS1 (cell proliferation, metastasis)	N/A	[107]
PRNCR1	N/A	Contains CRC-related SNPSs with its locus	N/A	[108]
PCAT-1	Up	N/A	N/A	[109]
PVT-1	Up	Suppresses apoptosis	TGF-β	[86,110]
uc73a	Up	Suppresses apoptosis	N/A	[111,112]
RAMS11	Up	Bind to CBX4 to regulate TOP2α	N/A	[113]
EPB41L4A-AS1	Up	Promote tumor cell proliferation, invasion, and migration	RhoA/ROCK signaling	[114]
Lnc-GNAT1-1	Down	Tumor suppressor	NF-κB signaling pathway	[115]
LUCAT-1	Up	Suppress tumor cell cycle and apoptosis	RPL40-MDM2-p53	[116]
cCSC1	Up	Increase cancer tumor cell self-renewal effect.	Hedgehog signaling pathway	[117]
HOXD-AS1	Down	Suppress tumor cell proliferation and migration	MAPK/AKT	[118]
DSCAM-AS1	Up	Promote tumor cell proliferation and migration	Notch signaling pathway	[90]
LNCAROD	Up	Tumor-educated platelets related	N/A	[119]
SNHG20	Up	Tumor-educated platelets related	N/A	[119]
LINC00534	Up	Tumor-educated platelets related	N/A	[119]
TSPOAP-AS1	Up	Tumor-educated platelets related	N/A	[119]

Table 1. Summary of potential lncRNA implicated in colorectal cancer, including dysregulation, potential function and mechanism, signaling pathways.

LncRNA	Dysregulation	Potential Function and Mechanism	Signaling Pathways	References
CCAL	Up	Suppress activator protein 2α and activate Wnt/ β-catenin pathway	Wnt/β-catenin	[95]
CCAT 1	Up	N/A	CCAT1/MYC	[96]
CCAT 2	Up	Stabilize and induce expression of BOP1 an activator of aurora kinase B	AURKB	[35]
CAHM	Down	N/A	N/A	[97]
CRNDE	Up	Suppress apoptosis	Wnt/β-catenin	[98,99,100]
H19	Up	mir675 regulate RB	N/A	[101]
HOTAIR	Up	Promote cell invasion, participates in EMT, metastasis	N/A	[102,103]
Loc285194	Down	Regulated by p53, inhibits tumor cell growth partially via repression of mir-211	N/A	[29,104]
LincRNA-p21	Down	Collaborates with hnRNP-K to act as a transcriptional coactivator of p53	N/A	[105]
LIT1/KCNQ1OT1	Loss of Imprinting	Associate with epigenetic status of KvDMR1	N/A	[84,106]
Lnc-LALC	Down	Inhibit expression of LZTS1 (cell proliferation, metastasis)	N/A	[107]
PRNCR1	N/A	Contains CRC-related SNPSs with its locus	N/A	[108]
PCAT-1	Up	N/A	N/A	[109]
PVT-1	Up	Suppresses apoptosis	TGF-β	[86,110]
uc73a	Up	Suppresses apoptosis	N/A	[111,112]
RAMS11	Up	Bind to CBX4 to regulate TOP2α	N/A	[113]
EPB41L4A-AS1	Up	Promote tumor cell proliferation, invasion, and migration	RhoA/ROCK signaling	[114]
Lnc-GNAT1-1	Down	Tumor suppressor	NF-κB signaling pathway	[115]
LUCAT-1	Up	Suppress tumor cell cycle and apoptosis	RPL40-MDM2-p53	[116]
cCSC1	Up	Increase cancer tumor cell self-renewal effect.	Hedgehog signaling pathway	[117]
HOXD-AS1	Down	Suppress tumor cell proliferation and migration	MAPK/AKT	[118]
DSCAM-AS1	Up	Promote tumor cell proliferation and migration	Notch signaling pathway	[90]
LNCAROD	Up	Tumor-educated platelets related	N/A	[119]
SNHG20	Up	Tumor-educated platelets related	N/A	[119]
LINC00534	Up	Tumor-educated platelets related	N/A	[119]
TSPOAP-AS1	Up	Tumor-educated platelets related	N/A	[119]

8. MicroRNAs in CRC

MicroRNAs have demonstrated potential as diagnosis biomarkers for early diagnosis, prognosis, prediction, as well as novel therapeutic targets in CRC [120,121,122]. Michal et al. first identified the link between miRNAs and CRC development by discovering that the expression of miR-143 and miR-145 was downregulated in CRC [123]. Since then, numerous microRNAs have been identified as independent factors associated with CRC. For instance, MiR-21, which is one oncogene in CRC and significantly upregulated in the plasma of CRC patients compared to the controls, can be used as a biomarker for early screening of CRC [29,121]. Similarly, miR-29a and miR-92a expression in serum levels are significantly higher in serums of patients with advanced stages of CRC compared with heathy controls [29,121]. In contrast, expression levels of miR-760, miR-601, and miR-548c-5p in serums of colorectal neoplasia are significantly decreased compared to healthy controls [124,125]. In addition, the dynamic changes in serum level of miR-150-5p, which have been observed to significantly decrease one month post-operation in CRC patients, can serve as a real-time indicator for monitoring of treatment effectiveness [126].

Furthermore, chemotherapy resistance in advanced CRC is closely associated with specific changes in miRNA expression. Increased serum levels of miR-19a and miR-19b are linked to resistance to FOLFOX and XELOX treatment, respectively [127,128]. Conversely, decreased levels of miR-4772-3p, miR-1914, and miR-1915 predict poor responses to FOLFOX and XELOX, with miR-128-3p playing a key role in enhancing sensitivity to oxaliplatin when overexpressed [128,129,130].

9. CircRNAs in CRC

CircRNAs are also abundant and stable in exosomes and can be detected in circulation and urine, and have been found to distinguish cancer patients from healthy controls [127,131,132,133,134]. These properties, combined with their unique expression profiles, allow circRNAs to serve as promising biomarkers for distinguishing cancer patients from healthy individuals and for prognostic assessment in CRC. Circ_KLDHC10, circ_0082182 and circ_0000370, circVAPA, and circ_002144 are significantly elevated, while plasma levels of circ_CCDC66, circ_ABCC1, circ_STIL, and circ_0035445 are markedly lower in CRC compared to healthy individuals [86,135,136,137]. Additionally, circ_0004771 is upregulated in serum exosomes of CRC patients but decreases significantly following surgery [138].

Similarly, low plasma levels of circ-ABCC1 and high plasma levels of circ_0004771 are associated with advanced stages of CRC and metastasis [135,136]. In addition, circ_0082182 and circ_0035445 demonstrated distinct expression patterns between preoperative and postoperative plasma in CRC patients [136]. Circ-0000338 was significantly downregulated in the serum exosomes of chemotherapy-resistant patients, indicating its potential as a biomarker for CRC resistance [139].

10. Method for Detecting ncRNAs

Various techniques are currently available for the detection and quantification of lncRNAs, microRNAs, and circRNAs. For lnRNAs, a common approach to identifying lncRNAs with prognostic potential in CRC involves expression profiling of tumors and healthy tissues using RNA sequencing or microarray, and these methods provide high-throughput data, while quantitative real-time PCR (qRT-PCR) is typically employed to validate the findings and ensure reliable gene expression results [85,140,141]. In addition, fluorescence in situ hybridization (FISH), which detects and quantifies target nucleic acid sequence by binding to complementary probes, is commonly used to assess the presence and abundance of specific lncRNAs [140,141]. More advanced methods for identifying lncRNA include RNA interference (RNAi), RNA pull-down assays, RNA-binding protein immunoprecipitation (RIP), chromatin isolation by RNA purification (ChIRP), ChIRP sequencing (ChIRP-seq), crosslinking immunoprecipitation (CLIP), and CLIP sequencing (CLIP-seq), etc. [140,142]. Similarly, microarray and qRT-PCR for targeted analysis of specific miRNAs and circRNA, next-generation sequencing (NGS), including RNA sequencing (RNA-seq), is used for comprehensive profiling of miRNA and circRNA expression across the whole transcriptome [143,144].

11. Tumor-Educated Platelets

Platelets, derived from megakaryocytes, play a crucial role in both hemostasis and cancer progression. In cancer patients, platelets interact closely with CTCs, supporting their survival, immune evasion, facilitate their seeding at secondary sites, and enhance outgrowth [14,145,146,147]. Specifically, platelets can recognize and interact with tumor cells, forming aggregates with CTCs and further expanding those aggregates through releasing prothrombotic and procoagulant microparticles or the expression of tissue factor in the circulation, which shield them from mechanical stress and immune attacks [14,146]. For example, platelet-released mediators, such as TGF-β, have been shown to accelerate epithelial–mesenchymal transition (EMT) in CTCs, thereby enhancing their invasive and metastasis [148]. In addition, platelets establish themselves as one of the richest sources of liquid biopsy through their ability to take up proteins and nucleic acids, and modify their megakaryocyte-derived transcripts and proteins in response to external signals, a process that gives rise to tumor-educated platelets (TEPs) with distinct protein and RNA profiles [14,149,150,151]. In CRC, a growing body of evidence demonstrates the potential of tumor-educated platelets (TEPs) for early detection. In a retrospective cohort study, Xu et al. analyzed platelets isolated from 132 patients with early and advanced CRC and 190 controls, where they demonstrated that the RNA profile of TEPs could serve as a diagnostic marker for identifying early-stage CRC, even in the presence of non-cancerous diseases. Furthermore, the study demonstrated high sensitivity and specificity in distinguishing between stages of CRC [14,18].

Peterson et al. found that vascular growth factor (VEGF), platelet factor 4 (PE4), and platelet-derived growth factor (PDGF) are elevated in platelets of 35 colorectal cancer patients compared to 84 healthy individuals [151]. Ye et al. discovered four platelet-related lncRNA-LNCAROD, SNHG20, LINC00534, and TSPOAP-AS1 are significantly upregulated in the platelets and serums of CRC patients, highlighting the potential of TEPs as biomarkers for monitoring the CRC progression [119]. Emerging evidence has revealed that TEPs activate coagulation cascade via tissue factor-mediated platelet activation, resulting in platelet-rich clots forming around CTCs. Those clots help protect CTCs from immune system clearance and enhance their survival [14]. Seretis C et al. demonstrated increased expression of invasion-related genes, including MYC, IL33, PTGS2, PTGER2, and VEGFB, suggesting that targeted therapies aimed at inhibiting platelet activity could potentially reduce tumor metastasis [152,153]. Li et al. discovered that elevated TEPs can contribute the growth and metastasis of CRC by binding CD62P to PSG-1 on tumor associated macrophages (TAMs), promoting the C5 transcription in TAMSs and triggering the C5a/C5Ar1 axis via JNK/STAT1 pathway [14,149]. Platelet-to-lymphocyte ratio (PLR) could also utilized to predict the response to adjuvant chemotherapy in stage II CRC patients, as a significant correlation has been observed between the PLR levels and the effectiveness of chemotherapy [154].

12. Method for Detecting TEPs

The primary traditional methods for detecting tumor-educated platelets (TEPs) include RNA-sequencing (RNA-seq) technology and reverse transcription–polymerase chain reaction (RT-PCR) [14]. In 2015, Best et al. introduced ThromboSeq platform, an innovative tool utilizing next-generation RNA-seq to identify TEP biomarkers for clinical applications [155]. Based on this foundation, their 2017 study demonstrated that the effectiveness of particle swarm optimization (PSO)-enhanced algorithms in selecting 799 RNA biomarker groups from platelet RNA-seq libraries [156]. This breakthrough enabled precise detection of both early and advanced non-small cell lung cancer based on TEP mRNA sequencing [156].

Further advancements were achieved in 2019 when PSO-enhanced algorithms were refined to automatically identify RNA sequences contributing more significantly to cancer classification. This innovation improved RNA-sequencing capabilities and enhanced cancer diagnostic accuracy [157]. In 2021, the team introduced imPlatelet classifiers, which transformed RNA-seq data from TEPs into visual images, with each pixel representing the expression level of a gene. Leveraging deep learning technology, this approach allowed for accurate cancer detection even with limited sample sizes [158]. In addition to mRNA, human platelets contain a variety of ncRNA, including lncRNA and circRNA, which are detected by RT-PCR, droplet digital PCR, and single-cell sequencing [156,157].

However, the standardization of TEP detection techniques remains a critical challenge. Factors such as pre-analytical and analytical processes require uniform protocols to enhance reliability and reproducibility. Addressing these issues is crucial for advancing TEP-based diagnostics and integrating them effectively into clinical practice.

13. Conclusions and Future Perspectives

The advent of liquid biopsy has revolutionized cancer diagnostics and disease monitoring, offering a minimally invasive alternative to traditional tissue biopsies. In colorectal cancer (CRC), liquid biopsy biomarkers, including circulating tumor cells (CTCs), circulating tumor DNA (ctDNA), long non-coding RNAs (lncRNAs), and tumor-educated platelets (TEPs) have shown immense potential in capturing the dynamic nature of tumor evolution. These biomarkers not only enable screening, early diagnosis, detection of MRD, real-time monitoring of treatment efficacy, and disease progression prediction, but also provide valuable insights into tumor heterogeneity and molecular alterations, guiding therapeutic strategies by identifying resistance mechanisms (e.g., KRAS mutations) and evaluating the efficacy of immunotherapies and targeted treatments [38]. For example, while CTCs primarily represent metastasis-initiating cells, ctDNA offers insights into mutations and genomic alterations across the entire tumor genome, ncRNAs highlight regulatory changes, making it highly associated with cancer severity and holds promise as a prognostic marker for the early detection of CRC recurrence [25,159].

However, despite its promise, challenges such as limited sensitivity, lack of standardization, and high costs remain significant barriers to its widespread adoption into routine clinical practice. A primary hurdle is detection sensitivity and specificity, particularly in early-stage CRC. For example, ctDNA detection rates in non-metastatic CRC can be as low as 50%, making early diagnosis challenging [37,57]. Additionally, variability in sample handling and pre-analytical conditions impact the accuracy and reproducibility of results, further complicating clinical implementation.

The standardization issue represents another significant barrier. The lack of uniform protocols for isolating, quantifying, and analyzing biomarkers has led to inconsistencies in sensitivity and specificity across different methodologies, such as PCR-based and NGS-based assays.

Cost and accessibility present additional limitations. Advanced detection technologies, such as NGS, require specialized infrastructure and expertise, making them less feasible in resource-constrained settings. The high cost of these technologies further restricts their routine clinical use.

Lastly, tumor heterogeneity adds complexity to liquid biopsy. While it provides systemic overview of tumor characteristics, it may not fully capture spatial variations within primary and metastatic sites. This limitation is particularly relevant for identifying minor subclones that contribute to the therapy resistance and disease progression.

To address these challenges and unlock the full potential of liquid biopsy, several opportunities for future advancements have been identified. One promising strategy is the development of multi-biomarker panels that combine CTCs, ctDNA, ncRNAs, and TEPs. Such integrated panels could overcome the limitations of individual methods and provide a more comprehensive and accurate approach to diagnosis, prognosis, and treatment monitoring. These uniform panels could significantly enhance the sensitivity and specificity of liquid biopsy, enabling comprehensive tumor profiling and facilitating personalized treatment strategies.

Technological advancements are also critical. Innovations like single-cell sequencing for CTCs and ultra-sensitive ctDNA assays hold the potential to improve early detection and monitoring capabilities. These technologies could provide more accurate and detailed insights into tumor biology, even in early-stage cancers.

Standardization is another critical priority for advancing the clinical utility of liquid biopsy. The lack of consistent methodologies for biomarker isolation, coupled with variability in sample handling, has led to diverse sensitivity and specificities, impacting the accuracy of the interpretation of testing results. Therefore, to address this, the development of standardized protocols for assays involving CTCs, ctDNA, ncRNA, TEPs is essential. Establishing uniform guidelines for biomarkers isolation, analysis, and interpretation will enhance reproducibility across institutions and studies, ensuring more reliable outcomes and facilitating the incorporation of liquid biopsy testing into routine clinical practice. Harmonizing guidelines will be essential for integrating liquid biopsy into routine clinical practice for CRC management on a global scale.

The integration of artificial intelligence (AI) offers further opportunities. AI-driven data analysis could enhance the interpretation of liquid biopsy results, providing real-time, accurate predictions of disease progression, treatment response, and resistance mechanisms. This would not only improve diagnostic accuracy but also optimize therapeutic decisions.