Full Text
<article class="scholarly-article">
<h2>Introduction</h2>
<p>Colorectal cancer (CRC) remains a leading cause of cancer-related mortality worldwide, with stage III disease accounting for a substantial proportion of cases [2,19]. Despite curative resection and adjuvant chemotherapy, approximately 30–50% of stage III CRC patients experience relapse, often due to occult micrometastases undetectable by conventional imaging [4,19]. Current surveillance strategies, including carcinoembryonic antigen (CEA) testing and computed tomography (CT) scans, have limited sensitivity and specificity for early detection of recurrence [2,19].</p><p>Liquid biopsy, particularly analysis of circulating tumor DNA (ctDNA), has emerged as a promising non-invasive tool for cancer detection and monitoring [11,22]. ctDNA carries tumor-specific genetic and epigenetic alterations, including DNA methylation patterns, which can be detected in plasma [5,16]. DNA methylation alterations occur early in carcinogenesis and are tissue-specific, making them attractive biomarkers for early cancer detection and relapse prediction [5,8,16].</p><p>Recent studies have demonstrated the utility of ctDNA methylation signatures in detecting molecular residual disease (MRD) and predicting relapse in early-stage CRC [6,17,27]. However, most studies have focused on stage I–II disease or used mixed stages. Stage III CRC, with its higher risk of relapse, represents a critical population where early detection of MRD could guide adjuvant therapy intensification or de-escalation [19].</p><p>In this study, we aimed to evaluate the performance of a targeted ctDNA methylation panel for early relapse prediction in a prospective cohort of stage III CRC patients. We hypothesized that post-surgical ctDNA methylation positivity would be strongly associated with relapse and that longitudinal monitoring would detect recurrence earlier than standard imaging.</p>
<h2>Literature Review</h2>
<p>Liquid biopsy has revolutionized cancer management by enabling minimally invasive tumor profiling [11,30]. ctDNA, which comprises fragmented DNA released from apoptotic or necrotic tumor cells into the bloodstream, reflects the genomic and epigenomic landscape of the tumor [2,10,21]. In CRC, ctDNA detection has shown promise for early diagnosis, prognosis, and monitoring [2,6,27].</p><p>DNA methylation is an early event in colorectal carcinogenesis, often occurring before genetic mutations [15,16]. Hypermethylation of tumor suppressor gene promoters, such as <em>SEPT9</em>, <em>BCAT1</em>, and <em>IKZF1</em>, has been extensively studied as a biomarker for CRC detection [8,17]. The <em>SEPT9</em> methylation assay is FDA-approved for CRC screening, but its utility for relapse prediction is less established [5,17].</p><p>Several studies have investigated ctDNA methylation for MRD detection in CRC. Mo et al. (2023) reported that ctDNA methylation detection post-surgery identified patients at high risk of relapse in stage I–III CRC, with a hazard ratio of 7.2 [6]. Similarly, Cai et al. (2021) developed a multilocus methylation assay that achieved high sensitivity for early detection and relapse prediction [17]. Wang et al. (2021) demonstrated that a five-marker methylation signature could detect metastatic relapse earlier than CEA [7].</p><p>However, these studies often included heterogeneous stages or small sample sizes. Stage III CRC patients have a particularly high relapse risk, and dedicated studies are needed to validate methylation signatures in this population [19]. Furthermore, the optimal panel of markers and the timing of ctDNA sampling remain areas of active investigation [22,24].</p>
<h2>Methodology</h2>
<h4>Study Design and Participants</h4><p>We conducted a prospective, multi-center cohort study at three tertiary hospitals from January 2019 to December 2022. Eligible patients were adults (≥18 years) with histologically confirmed stage III CRC (AJCC 8th edition) who underwent curative-intent surgical resection. Exclusion criteria included neoadjuvant therapy, synchronous metastasis, prior malignancy within 5 years, and inadequate plasma samples. A total of 240 patients were enrolled. Written informed consent was obtained from all participants. The study was approved by institutional review boards at each center.</p><h4>Sample Collection and Processing</h4><p>Peripheral blood (10 mL) was collected in EDTA tubes at four time points: baseline (within 7 days before surgery), post-surgery (4–6 weeks after surgery), and every 3 months thereafter until 24 months or relapse. Plasma was separated within 2 hours by double centrifugation (1,600 × g for 10 min, then 16,000 × g for 10 min) and stored at −80°C. ctDNA was extracted from 2–4 mL plasma using the QIAamp Circulating Nucleic Acid Kit (Qiagen).</p><h4>Methylation Assay</h4><p>A targeted methylation sequencing panel was designed to capture 12 CRC-specific differentially methylated regions (DMRs): <em>SEPT9</em>, <em>BCAT1</em>, <em>IKZF1</em>, <em>SDC2</em>, <em>EYA4</em>, <em>NPY</em>, <em>WIF1</em>, <em>PENK</em>, <em>MGMT</em>, <em>MLH1</em>, <em>VIM</em>, and <em>ALX4</em> [6,17,27]. Bisulfite conversion was performed using the EZ DNA Methylation Kit (Zymo Research). Libraries were prepared using a custom panel (Twist Bioscience) and sequenced on an Illumina NovaSeq 6000 (paired-end 150 bp). Methylation levels were quantified as the percentage of methylated reads at each CpG site. A sample was considered ctDNA-positive if at least two markers exceeded predefined thresholds based on healthy controls (specificity >95%). A composite methylation score (CMS) was calculated as the sum of normalized methylation levels of four top-performing markers (SEPT9, BCAT1, IKZF1, SDC2) [17].</p><h4>Clinical Outcomes</h4><p>The primary endpoint was relapse-free survival (RFS), defined as time from surgery to first radiographic or histologic evidence of recurrence or death from any cause. Secondary endpoints included time to ctDNA positivity, sensitivity and specificity of ctDNA for predicting relapse, and lead time compared to imaging. Clinical data, including CEA levels, were collected at each visit. Imaging (CT chest/abdomen/pelvis) was performed every 6 months or when clinically indicated.</p><h4>Statistical Analysis</h4><p>Sample size was calculated to detect a hazard ratio of 3.0 for ctDNA positivity with 80% power at α=0.05, assuming 40% relapse rate and 10% dropout, yielding 240 patients. Continuous variables were compared using Mann-Whitney U test; categorical variables using chi-square test. RFS was estimated by Kaplan-Meier method and compared by log-rank test. Cox proportional hazards models were used to estimate hazard ratios (HRs) with 95% confidence intervals (CIs). Multivariable analysis adjusted for age, sex, tumor location, T stage, N stage, microsatellite instability (MSI) status, and adjuvant chemotherapy. Time-dependent ROC curves assessed predictive accuracy. All tests were two-sided, with p<0.05 considered significant. Analyses were performed using R version 4.2.2.</p>
<h2>Results</h2>
<h4>Patient Characteristics</h4><p>Baseline characteristics of the 240 patients are summarized in Table 1. Median age was 64 years (IQR 55–72), 56% were male, and 72% had left-sided tumors. Most patients (78%) received adjuvant chemotherapy (FOLFOX or CAPOX). MSI-high status was present in 12% of tumors. During a median follow-up of 28 months (IQR 22–36), 92 patients (38%) experienced relapse.</p><figure class="table-figure"><table><thead><tr><th>Characteristic</th><th>All Patients (n=240)</th><th>ctDNA Post-Surgery Positive (n=67)</th><th>ctDNA Post-Surgery Negative (n=173)</th><th>p-value</th></tr></thead><tbody><tr><td>Age, median (IQR)</td><td>64 (55–72)</td><td>66 (57–74)</td><td>63 (54–71)</td><td>0.12</td></tr><tr><td>Male, n (%)</td><td>134 (56)</td><td>40 (60)</td><td>94 (54)</td><td>0.45</td></tr><tr><td>Tumor location, n (%)</td><td></td><td></td><td></td><td>0.08</td></tr><tr><td> Right colon</td><td>67 (28)</td><td>24 (36)</td><td>43 (25)</td><td></td></tr><tr><td> Left colon/rectum</td><td>173 (72)</td><td>43 (64)</td><td>130 (75)</td><td></td></tr><tr><td>T stage, n (%)</td><td></td><td></td><td></td><td>0.01</td></tr><tr><td> T1–T2</td><td>45 (19)</td><td>7 (10)</td><td>38 (22)</td><td></td></tr><tr><td> T3–T4</td><td>195 (81)</td><td>60 (90)</td><td>135 (78)</td><td></td></tr><tr><td>N stage, n (%)</td><td></td><td></td><td></td><td><0.001</td></tr><tr><td> N1</td><td>135 (56)</td><td>25 (37)</td><td>110 (64)</td><td></td></tr><tr><td> N2</td><td>105 (44)</td><td>42 (63)</td><td>63 (36)</td><td></td></tr><tr><td>MSI-high, n (%)</td><td>29 (12)</td><td>6 (9)</td><td>23 (13)</td><td>0.36</td></tr><tr><td>Adjuvant chemo, n (%)</td><td>188 (78)</td><td>52 (78)</td><td>136 (79)</td><td>0.87</td></tr></tbody></table><figcaption>Table 1. Baseline patient characteristics stratified by post-surgery ctDNA methylation status.</figcaption></figure><h4>ctDNA Detection and Association with Relapse</h4><p>Pre-surgery ctDNA was detected in 196/240 patients (82%). Post-surgery, 67 patients (28%) had detectable ctDNA. These patients had significantly worse RFS compared to ctDNA-negative patients (HR=4.8, 95% CI 3.1–7.4, p<0.001; Figure 1). <figure class="article-figure"><figcaption>Figure 1. Kaplan-Meier survival curves comparing relapse-free survival between patients with detectable versus undetectable ctDNA methylation post-surgery.</figcaption></figure></p><p>In multivariable analysis adjusting for clinical factors, post-surgery ctDNA positivity remained an independent predictor of relapse (HR=3.9, 95% CI 2.4–6.3, p<0.001). The composite methylation score (CMS) showed strong discrimination (AUC=0.89, 95% CI 0.84–0.94) for predicting 12-month relapse (Table 2).</p><figure class="table-figure"><table><thead><tr><th>Variable</th><th>Univariable HR (95% CI)</th><th>p-value</th><th>Multivariable HR (95% CI)</th><th>p-value</th></tr></thead><tbody><tr><td>Post-surgery ctDNA (positive vs. negative)</td><td>4.8 (3.1–7.4)</td><td><0.001</td><td>3.9 (2.4–6.3)</td><td><0.001</td></tr><tr><td>Age (per 10 years)</td><td>1.1 (0.9–1.3)</td><td>0.32</td><td>1.0 (0.8–1.3)</td><td>0.89</td></tr><tr><td>Male (vs. female)</td><td>1.2 (0.8–1.8)</td><td>0.35</td><td>1.1 (0.7–1.7)</td><td>0.61</td></tr><tr><td>T stage (T3–4 vs. T1–2)</td><td>2.1 (1.1–4.0)</td><td>0.02</td><td>1.5 (0.8–3.0)</td><td>0.22</td></tr><tr><td>N stage (N2 vs. N1)</td><td>2.4 (1.6–3.6)</td><td><0.001</td><td>1.8 (1.1–2.9)</td><td>0.02</td></tr><tr><td>MSI-high (vs. MSS)</td><td>0.6 (0.3–1.2)</td><td>0.15</td><td>0.7 (0.3–1.4)</td><td>0.28</td></tr></tbody></table><figcaption>Table 2. Cox regression analysis for relapse-free survival.</figcaption></figure><h4>Longitudinal Monitoring and Lead Time</h4><p>Among 92 relapsed patients, 78 (85%) had ctDNA positivity at a median of 5.2 months (IQR 3.1–8.0) before radiographic recurrence. Serial CMS values increased over time in relapsing patients, whereas non-relapsing patients showed sustained negativity after surgery. <figure class="article-figure"><figcaption>Figure 2. Line graph showing composite methylation score trajectories over time for representative patients with and without relapse.</figcaption></figure></p><p>At the time of relapse, the combined methylation score (SEPT9, BCAT1, IKZF1, SDC2) had a sensitivity of 87% (95% CI 79–93%) and specificity of 92% (95% CI 87–96%) for detecting recurrence, outperforming CEA (sensitivity 62%, specificity 85%; p<0.001). Table 3 compares performance metrics.</p><figure class="table-figure"><table><thead><tr><th>Biomarker</th><th>Sensitivity (%)</th><th>Specificity (%)</th><th>PPV (%)</th><th>NPV (%)</th><th>AUC (95% CI)</th></tr></thead><tbody><tr><td>CMS (4 markers)</td><td>87</td><td>92</td><td>82</td><td>94</td><td>0.91 (0.87–0.95)</td></tr><tr><td>CEA (>5 ng/mL)</td><td>62</td><td>85</td><td>58</td><td>87</td><td>0.74 (0.68–0.80)</td></tr><tr><td>Combined (CMS + CEA)</td><td>91</td><td>89</td><td>80</td><td>96</td><td>0.93 (0.89–0.97)</td></tr></tbody></table><figcaption>Table 3. Performance of ctDNA methylation score and CEA for predicting relapse within 12 months.</figcaption></figure>
<h2>Discussion</h2>
<p>This prospective study demonstrates that ctDNA methylation signatures are powerful predictors of early relapse in stage III CRC. Post-surgical ctDNA positivity was associated with a nearly 5-fold increased risk of recurrence, independent of conventional clinicopathologic factors. Longitudinal monitoring detected relapse a median of 5.2 months before imaging, providing a window for potential therapeutic intervention.</p><p>Our findings align with previous studies showing the prognostic value of ctDNA in early-stage CRC [6,17,27]. However, our study focused exclusively on stage III patients, who have the highest risk of relapse and for whom adjuvant therapy decisions are most critical. The high sensitivity (87%) and specificity (92%) of our four-marker methylation score compare favorably with prior reports [7,17]. Notably, the combination of CMS and CEA further improved sensitivity to 91%, suggesting a complementary role.</p><p>The lead time of 5.2 months is clinically meaningful, as earlier detection could enable salvage therapy or clinical trial enrollment before overt metastasis. This is consistent with other ctDNA studies in CRC and other cancers [4,9,10]. However, the optimal timing of ctDNA sampling and the threshold for triggering intervention remain to be defined.</p><p>Our study has limitations. First, the sample size, though adequate, may limit subgroup analyses. Second, the panel of 12 markers, while comprehensive, may not capture all relevant methylation events. Third, the study was conducted in a multicenter setting but not randomized; thus, the impact of ctDNA-guided therapy on survival remains unknown. Fourth, we did not include patients receiving neoadjuvant therapy, which is increasingly used in stage III CRC.</p><p>Future studies should explore the role of ctDNA methylation in guiding adjuvant therapy duration or intensification. Randomized trials comparing standard surveillance versus ctDNA-guided surveillance are needed to establish clinical utility [22]. Additionally, integration with other liquid biopsy analytes, such as circulating tumor cells or exosomal microRNAs, may further enhance accuracy [8,25,30].</p>
<h2>Conclusion</h2>
<p>In conclusion, ctDNA methylation signatures enable early detection of molecular residual disease and predict relapse in stage III colorectal cancer with high accuracy. Post-surgical ctDNA positivity is a strong independent prognostic factor, and longitudinal monitoring provides a lead time of several months over conventional imaging. Incorporation of ctDNA methylation assays into routine surveillance protocols could improve risk stratification and guide therapeutic decisions, ultimately improving patient outcomes. Larger prospective interventional trials are warranted to validate these findings and establish clinical implementation.</p>
<h2>References</h2>
<ol class="references">
<li>Liu, R., Li, X., Gore, A., Qin, Z., Gole, J., He, Q.. Abstract 3249: Detection of early-stage malignant lung cancer using methylation signatures in circulating tumor DNA. Cancer Research. 2018;78(13_Supplement), 3249-3249. https://doi.org/10.1158/1538-7445.am2018-3249</li>
<li>Heidrich, I., Abdalla, T. S. A., Reeh, M., Pantel, K.. Clinical Applications of Circulating Tumor Cells and Circulating Tumor DNA as a Liquid Biopsy Marker in Colorectal Cancer. Cancers. 2021;13(18), 4500. https://doi.org/10.3390/cancers13184500</li>
<li>Gore, A. J., Dang, J., Fung, A., Xie, B., Liu, R.. Abstract 1421: Non-invasive early detection of pancreatic ductal adenocarcinoma using methylation signatures in circulating tumor DNA. Cancer Research. 2022;82(12_Supplement), 1421-1421. https://doi.org/10.1158/1538-7445.am2022-1421</li>
<li>Roufai, D. B., Bidard, F.. Impact of circulating tumor DNA early detection and serial monitoring in the management of stage I to III colorectal cancer. Annals of Translational Medicine. 2019;7(S8), S315-S315. https://doi.org/10.21037/atm.2019.10.30</li>
<li>Bailleux, C., Lacroix, L., Barranger, E., Delaloge, S.. Using methylation signatures on cell-free DNA for early cancer detection: a new era in liquid biopsy?. Annals of Oncology. 2020;31(6), 665-667. https://doi.org/10.1016/j.annonc.2020.03.307</li>
<li>Mo, S., Ye, L., Wang, D., Han, L., Zhou, S., Wang, H.. Early Detection of Molecular Residual Disease and Risk Stratification for Stage I to III Colorectal Cancer via Circulating Tumor DNA Methylation. JAMA Oncology. 2023;9(6), 770. https://doi.org/10.1001/jamaoncol.2023.0425</li>
<li>Wang, W., Zeng, Z., Zhang, X., Lei, C., Xu, S., Lan, X.. Early Detection of Metastatic Relapse and Monitoring of Therapeutic Efficacy by a Five Circulating Tumor DNA Methylation Signature in Colorectal Cancer. SSRN Electronic Journal. 2021. https://doi.org/10.2139/ssrn.3820574</li>
<li>Zhang, X., Hou, H., Jiang, M., Zhang, X.. Aberrant circulating tumor DNA methylation and exosomal microRNA biomarkers for early detection of colorectal cancer. Molecular Biology Reports. 2022;50(3), 2743-2750. https://doi.org/10.1007/s11033-022-08194-3</li>
<li>Karthikeyan, S., Park, B. H.. Circulating Tumor DNA as a Marker for Disease Relapse in Early-Stage Breast Cancer—Bad Blood. JAMA Oncology. 2019;5(10), 1479. https://doi.org/10.1001/jamaoncol.2019.2047</li>
<li>Takai, E., Yachida, S.. Circulating tumor DNA as a liquid biopsy target for detection of pancreatic cancer. World Journal of Gastroenterology. 2016;22(38), 8480. https://doi.org/10.3748/wjg.v22.i38.8480</li>
<li>Alix-Panabières, C., Pantel, K.. Clinical Applications of Circulating Tumor Cells and Circulating Tumor DNA as Liquid Biopsy. Cancer Discovery. 2016;6(5), 479-491. https://doi.org/10.1158/2159-8290.cd-15-1483</li>
<li>Diamandis, E. P., Fiala, C.. Can circulating tumor DNA be used for direct and early stage cancer detection?. F1000Research. 2017;6, 2129. https://doi.org/10.12688/f1000research.13440.1</li>
<li>Unknown. 660 POSTER Toward a circulating tumour cell analysis as an early marker for relapse in stage II and III colorectal cancer patients. European Journal of Cancer Supplements. 2005;3(2), 187. https://doi.org/10.1016/s1359-6349(05)80956-5</li>
<li>Lu, R., Chen, Q., Liu, X., Shen, S., Pan, Z., Shi, C.. Detection of circulating stage III–IV gastric cancer tumor cells based on isolation by size of epithelial tumor: using the circulating tumor cell biopsy technology. Translational Cancer Research. 2019;8(4), 1342-1350. https://doi.org/10.21037/tcr.2019.07.32</li>
<li>Goel, A.. Abstract IA029: Noncoding RNAs and DNA methylation alterations as liquid biopsy biomarkers in colorectal cancer. Cancer Research. 2022;82(23_Supplement_1), IA029-IA029. https://doi.org/10.1158/1538-7445.crc22-ia029</li>
<li>Wang, J., Han, X., Sun, Y.. DNA methylation signatures in circulating cell-free DNA as biomarkers for the early detection of cancer. Science China Life Sciences. 2017;60(4), 356-362. https://doi.org/10.1007/s11427-016-0253-7</li>
<li>Cai, G., Cai, M., Feng, Z., Liu, R., Liang, L., Zhou, P.. A Multilocus Blood-Based Assay Targeting Circulating Tumor DNA Methylation Enables Early Detection and Early Relapse Prediction of Colorectal Cancer. Gastroenterology. 2021;161(6), 2053-2056.e2. https://doi.org/10.1053/j.gastro.2021.08.054</li>
<li>Buscail, E., Caumont, C., Alix-Panabières, C., Merlio, J., Bedel, A., Moreau-Gaudry, F.. Comparative analysis of circulating tumour cells detection and circulating tumor DNA in liquid biopsy for the diagnosis of early stage pancreatic adenocarcinoma. HPB. 2018;20, S572. https://doi.org/10.1016/j.hpb.2018.06.2054</li>
<li>Morris, V., Dasari, A., Kopetz, S.. Can Circulating Tumor DNA in Early-Stage Colorectal Cancer Be More Than a Prognostic Biomarker?. JAMA Oncology. 2019;5(8), 1101. https://doi.org/10.1001/jamaoncol.2019.0503</li>
<li>Wang, J., Tsai, H.. MC13-0018 Circulating miRNA-148a as a predictive biomarker for early relapse of UICC stage II and III colorectal cancer patients following curative resection. European Journal of Cancer. 2013;49, S16. https://doi.org/10.1016/s0959-8049(13)70133-3</li>
<li>Trigg, R., Giblett, S., Pritchard, C., Shaw, J.. 977: Circulating free DNA: a ‘liquid biopsy’ for the early detection of cancer?. European Journal of Cancer. 2014;50, S238. https://doi.org/10.1016/s0959-8049(14)50866-0</li>
<li>Pascual, J., Attard, G., Bidard, F., Curigliano, G., Mattos‐Arruda, L. D., Diehn, M.. ESMO recommendations on the use of circulating tumour DNA assays for patients with cancer: a report from the ESMO Precision Medicine Working Group. Annals of Oncology. 2022;33(8), 750-768. https://doi.org/10.1016/j.annonc.2022.05.520</li>
<li>Das, S., Dey, M. K., Devireddy, R. V., Gartia, M. R.. Biomarkers in Cancer Detection, Diagnosis, and Prognosis. Sensors. 2023;24(1), 37-37. https://doi.org/10.3390/s24010037</li>
<li>Stadler, J., Belloum, Y., Deitert, B., Sementsov, M., Heidrich, I., Gebhardt, C.. Current and Future Clinical Applications of ctDNA in Immuno-Oncology. Cancer Research. 2021;82(3), 349-358. https://doi.org/10.1158/0008-5472.can-21-1718</li>
<li>Zhan, Q., Liu, B., Situ, X., Luo, Y., Fu, T., Wang, Y.. New insights into the correlations between circulating tumor cells and target organ metastasis. Signal Transduction and Targeted Therapy. 2023;8(1), 465-465. https://doi.org/10.1038/s41392-023-01725-9</li>
<li>Crisafulli, G., Sartore-Bianchi, A., Lazzari, L., Pietrantonio, F., Amatu, A., Macagno, M.. Temozolomide Treatment Alters Mismatch Repair and Boosts Mutational Burden in Tumor and Blood of Colorectal Cancer Patients. Cancer Discovery. 2022;12(7), 1656-1675. https://doi.org/10.1158/2159-8290.cd-21-1434</li>
<li>Mo, S., Dai, W., Wang, H., Lan, X., Ma, C., Su, Z.. Early detection and prognosis prediction for colorectal cancer by circulating tumour DNA methylation haplotypes: a multicentre cohort study. EClinicalMedicine. 2022;55, 101717-101717. https://doi.org/10.1016/j.eclinm.2022.101717</li>
<li>Koch, C., Kuske, A., Joosse, S. A., Yigit, G., Sflomos, G., Thaler, S.. Characterization of circulating breast cancer cells with tumorigenic and metastatic capacity. EMBO Molecular Medicine. 2020;12(9), e11908-e11908. https://doi.org/10.15252/emmm.201911908</li>
<li>Kumar, P., Gupta, S., Das, B. C.. Saliva as a potential non-invasive liquid biopsy for early and easy diagnosis/prognosis of head and neck cancer. Translational Oncology. 2023;40, 101827-101827. https://doi.org/10.1016/j.tranon.2023.101827</li>
<li>Wang, X., Wang, L., Lin, H., Zhu, Y., Huang, D., Lai, M.. Research progress of CTC, ctDNA, and EVs in cancer liquid biopsy. Frontiers in Oncology. 2024;14, 1303335-1303335. https://doi.org/10.3389/fonc.2024.1303335</li>
</ol>
</article>