Journal of Clinical Oncology July 29, 2026

Dynamic Circulating Tumor DNA Methylation Monitoring Guiding Postoperative Surveillance in Nonmetastatic Colorectal Cancer: A Prospective, Randomized, Phase III FIND Trial

Shaobo Mo, Chaoqiang Zhou, Maoguang Ma, Wenqin Luo, Yaqun Li ... Junjie Peng, et al.

Bottom Line

In a phase III randomized trial, dynamic ctDNA methylation-guided surveillance of nonmetastatic colorectal cancer detected recurrences earlier than standard imaging, doubling the proportion of patients eligible for curative-intent metastasis-directed therapy.

Key Findings

1. Overall recurrence rates were similar between the ctDNA-guided group and the standard surveillance control group (18.0% vs 18.6%, P = 0.919) at a median follow-up of 23.3 months.
2. The primary endpoint, the rate of curative-intent treatment for recurrence, was significantly higher in the ctDNA-guided group compared to the control group (48.1% vs 23.6%, relative risk [RR] 2.03, P = 0.008).
3. The median time to clinical recurrence was significantly shorter in the ctDNA-guided cohort (9.5 months vs 13.4 months, P < 0.001), demonstrating a diagnostic lead time of 3.9 months.
4. For recurrences confined to the liver and/or lungs, the ctDNA-guided group achieved a substantially higher curative resection rate (42.3% vs 18.2%, P = 0.002).
5. Hepatic recurrences in the ctDNA-guided group presented with more favorable and resectable features, including fewer lesions (≤3 lesions: 75.0% vs 28.6%, P = 0.005), smaller tumor size (≤3 cm: 90.0% vs 57.1%, P = 0.033), and more unilobar disease (80.0% vs 28.6%, P = 0.002).

Study Design

Design
RCT
Open-Label
Sample
584
Patients
Duration
23.3 mo
Median
Setting
Multicenter, China
Population Eligible patients with nonmetastatic colorectal cancer (CRC) who had undergone curative-intent surgical resection.
Intervention Dynamic surveillance guided by circulating tumor DNA (ctDNA) methylation testing, wherein a positive ctDNA result triggered immediate CT imaging. If negative, patients underwent bimonthly CT alongside quarterly ctDNA testing (reverting to standard imaging frequency after two consecutive negative ctDNA results).
Comparator Standard computed tomography (CT)-based monitoring.
Outcome The proportion of patients with disease recurrence who received curative-intent metastasis-directed therapy.

Study Limitations

The median follow-up of 23.3 months is relatively short; long-term data on disease-free survival (DFS) and overall survival (OS) are needed to determine if increased salvage resection rates translate to a definitive mortality benefit.
The open-label study design inherently risks introducing bias regarding the intensity of diagnostic follow-up and the subsequent surgical decision-making for recurrences.
Shorter time to clinical recurrence in the experimental arm inherently reflects lead-time bias, making survival from the point of recurrence a potentially confounded metric in future analyses.
The trial utilized a specific 10-marker ctDNA methylation assay, which may limit the direct generalizability of these findings to tumor-informed mutational assays or other liquid biopsy platforms.

Clinical Significance

The FIND trial establishes a new paradigm for postoperative surveillance in nonmetastatic colorectal cancer. By shifting from fixed-schedule radiological imaging to a dynamic, molecularly-guided approach, clinicians can identify recurrences at an earlier, oligometastatic stage. This nearly doubles the proportion of patients who can undergo potentially curative salvage surgery, transforming the management of minimal residual disease. If long-term follow-up confirms an overall survival benefit, ctDNA-guided surveillance will likely become a new standard of care for CRC monitoring.

Historical Context

Historically, standard surveillance (using CT scans and CEA levels) following the curative resection of nonmetastatic CRC has detected recurrences, but often too late; only about 20% to 25% of patients with detected recurrences have historically been eligible for curative-intent surgery. In recent years, circulating tumor DNA (ctDNA) emerged as a highly sensitive biomarker for minimal residual disease (MRD). While massive landmark studies (such as GALAXY, TRACC, and COBRA) primarily investigated ctDNA's role in guiding adjuvant chemotherapy decisions immediately post-surgery, the FIND trial is uniquely pivotal as a phase III study evaluating longitudinal ctDNA monitoring to explicitly dictate postoperative surveillance intensity and imaging schedules.

Guided Discussion

High-yield insights from every perspective

Med Student
Medical Student

What is the biological basis for using circulating tumor DNA (ctDNA) methylation patterns, rather than just genetic mutations, as a biomarker for minimal residual disease in colorectal cancer?

Key Response

Methylation changes (epigenetics) occur early in tumorigenesis and are often highly consistent across tumor types. They can offer a broader target space and higher clinical sensitivity than single-point mutations, making them excellent markers for detecting microscopic recurrence before anatomic changes are visible on traditional imaging.

Resident
Resident

In a patient with stage III colon cancer who recently completed adjuvant chemotherapy, how would the integration of ctDNA monitoring alter the standard postoperative surveillance algorithm, and what is the ultimate clinical goal of finding recurrence earlier?

Key Response

Standard surveillance relies on serial CEA and CT scans. Integrating ctDNA allows for 'molecular relapse' detection months before radiographic relapse. The primary clinical goal, as demonstrated by the FIND trial, is to identify oligometastatic disease (e.g., isolated liver metastases) earlier, significantly increasing the chance that the patient can undergo curative-intent surgical resection rather than just palliative systemic therapy.

Fellow
Fellow

While the FIND trial showed an increase in curative-intent metastasis-directed therapy via ctDNA monitoring, how should a positive ctDNA result with concurrent negative imaging be managed, and what are the risks of 'molecular anxiety' in these patients?

Key Response

Managing 'ctDNA-positive/scan-negative' patients is a major clinical dilemma. Currently, there is no consensus on whether to initiate systemic therapy blindly or wait for radiologic confirmation. Intervening too early might select for resistant clones without improving overall survival, while waiting causes significant psychological distress (molecular anxiety) for the patient knowing they have a recurrence that cannot yet be localized.

Attending
Attending

The FIND trial demonstrates a doubling of eligibility for curative-intent therapy; however, before universally adopting this in our practice, what ultimate endpoint must be proven, and how do we weigh the financial toxicity of dynamic ctDNA monitoring?

Key Response

While increased eligibility for curative intent is an excellent surrogate, the ultimate gold-standard endpoint is Overall Survival (OS). We must prove that earlier detection and surgery actually cure patients rather than just subjecting them to lead-time bias or morbidity from futile surgeries, especially given the high cumulative cost and financial toxicity of serial ctDNA assays.

Scholarly Review

Critical appraisal through the lens of expert reviewers and guideline development

PhD
PhD

How must a trial like FIND rigorously account for lead-time and length-time biases when evaluating whether earlier detection of recurrence by ctDNA actually translates to improved long-term oncologic outcomes?

Key Response

Lead-time bias occurs because ctDNA detects recurrence earlier than CT, artificially inflating survival time from detection without actually extending the patient's life. Length-time bias might occur if ctDNA preferentially detects slow-growing tumors. The trial must use Overall Survival from the time of initial randomization (not time of recurrence detection) as the primary or key secondary endpoint to definitively prove biological benefit.

Journal Editor
Journal Editor

As a reviewer evaluating the FIND trial, how would you scrutinize the definition of 'eligibility for curative-intent metastasis-directed therapy', and what risk of investigator bias exists in an unblinded study utilizing this as an endpoint?

Key Response

'Eligibility for curative therapy' can be a subjective endpoint heavily influenced by the multidisciplinary tumor board's aggressiveness. In an unblinded trial where investigators know a patient is in the experimental (ctDNA) arm, there is a high risk of performance and detection bias. Clinicians might look harder for resectable disease or push boundaries for surgery, which could artificially inflate the experimental arm's success rate.

Guideline Committee
Guideline Committee

Current NCCN and ESMO guidelines for CRC surveillance emphasize CEA and cross-sectional imaging but do not yet mandate routine ctDNA monitoring. Based on the Phase III FIND trial data, does this evidence meet the threshold to upgrade ctDNA from an 'emerging tool' to a standard-of-care recommendation for postoperative surveillance?

Key Response

While the Phase III FIND trial provides strong Level 1 evidence for detecting recurrences and improving curative-intent surgical rates, guideline committees usually require mature Overall Survival (OS) data before mandating a costly, paradigm-shifting surveillance strategy. The committee would likely issue a weak recommendation or state that ctDNA 'can be considered' in high-risk patients, pending long-term OS results.

Clinical Landscape

Noteworthy Related Trials

2014

FACS Trial

n = 1202 · JAMA

Tested

Intensive surveillance (CEA and/or CT)

Population

Patients with curatively resected colorectal cancer

Comparator

Minimum standard surveillance

Endpoint

Surgical treatment of recurrence with curative intent

Key result: Intensive surveillance increased the detection of curable recurrences but did not significantly improve overall survival compared to minimal follow-up.
2022

DYNAMIC Trial

n = 455 · NEJM

Tested

ctDNA-guided adjuvant therapy

Population

Patients with stage II colon cancer

Comparator

Standard clinicopathological management

Endpoint

2-year recurrence-free survival

Key result: A ctDNA-guided approach significantly reduced the use of adjuvant chemotherapy without compromising recurrence-free survival.
2023

GALAXY Study

n = 1039 · Nat Med

Tested

Postoperative ctDNA monitoring

Population

Patients with resected stage II-IV colorectal cancer

Comparator

ctDNA negative patients (observational)

Endpoint

Disease-free survival

Key result: Postoperative ctDNA positivity was highly predictive of recurrence, and clearance of ctDNA following adjuvant chemotherapy was associated with improved outcomes.

Tailored to your role

Want this tailored to you?

Add your specialty or training stage to get role-specific takeaways and more questions.

Personalize this analysis