ctDNA-Guided Surveillance Leads to Increased Curative-Intent Treatment After Recurrence in Nonmetastatic CRC

The use of circulating tumor DNA (ctDNA)–guided surveillance led to earlier recurrence detection and an increased use of curative-intent treatment after recurrence compared with standard surveillance in patients with nonmetastatic colorectal cancer (CRC), according to data from the phase 3 FIND trial (NCT05904665) presented at the 2025 ESMO Gastrointestinal Cancers Congress.

Findings demonstrated that in the ctDNA surveillance arm (n = 289) and the control arm (n = 295), the recurrence rates were 8.3% and 10.5% respectively. Among patients who experienced disease recurrence, 50% (95% CI, 31.4%-68.6%) of patients in the ctDNA arm (n = 24) received curative-intent treatment compared with 22.6% (95% CI, 11.4%-39.8%) of patients in the standard surveillance arm (n = 31; relative risk, 2.214; 95% CI, 1.060-4.780; P = .034).

“The FIND trial methodology demonstrates the feasibility of a ctDNA methylation assay with a clinically actionable turnaround time for reporting results to guide surveillance decisions [in nonmetastatic CRC],” lead study author Junjie Peng, MD, said in a presentation of the late-breaking abstract. Peng is chief physician and professor in the Department of Colorectal Surgery at Fudan University Shanghai Cancer Center in China.

What Investigators Hoped to FIND

Following surgical resection, surveillance is vital to detect recurrence of nonmetastatic CRC and to allow for timely intervention; however, Peng explained that prior data have shown that approximately 20% of patients who experience recurrence detected via surveillance have the opportunity for surgical treatment with curative intent.

To assess the feasibility of ctDNA monitoring as a surveillance tool, the prospective, randomized FIND trial enrolled patients with at least 18 years of age with nonmetastatic CRC who had an ECOG performance status of 0 to 2, were candidates for curative-intent surgery, and had a life expectancy of at least 12 months.

Patients were randomly assigned 1:1 between the two arms. In the control arm, patients underwent surgery and standard surveillance, which comprised physical examination and CEA testing; CT scans of the chest, abdomen, and pelvis; and colonoscopy, as per the National Comprehensive Cancer Network Guidelines.

In the experimental arm, ctDNA testing occurred within 1 month prior to surgery, then every 3 months for 2 years. If patients in this cohort tested positive for ctDNA, they underwent an immediate CT scan or MRI; if the scan was also positive, a multidisciplinary team (MDT) determined the next course of treatment. If the CT scan or MRI was negative, patients received 2 subsequent scans within the following 6 months. A positive scan would lead to an MDT discussion, and if both subsequent scans were also negative, the patient would return to the ctDNA monitoring protocol. The imaging process was triggered by any positive ctDNA test.

ctDNA testing was conducted using an improved single-tube, methylation-specific assay designed to detect 10 methylation markers with the ability to detect ctDNA down to a 0.01% allele fraction.

The primary end point was the proportion of patients receiving curative-intent therapy. Secondary end points included time to clinical recurrence (TTCR), disease-free survival, overall survival, and health-related quality of life.

Patient Population

Among 795 screened patients, 728 were randomly assigned between the 2 arms. In the 363 patients assigned to the ctDNA group, 74 did not undergo monitoring due to stage IV disease (n = 18), a diagnosis of advanced adenoma (n = 6), a non-CRC diagnosis (n = 9), and no surgery performed (n = 41). Of the 365 patients assigned to standard surveillance, 70 did not undergo postoperative follow-up due to stage IV disease (n = 16), a diagnosis of advanced adenoma (n = 8), a non-CRC diagnosis (n = 7), and no surgery performed (n = 39).

A total of 289 patients in the ctDNA arm and 295 patients in the surveillance arm were included in the intention-to-treat (ITT) population; 25 patients and 11 patients, respectively, were not included in the per-protocol population due to violations.

In the ITT population, patients ranged in age from 18 to 44 years (ctDNA arm, 6.9%; surveillance arm, 10.5%), 45 to 64 years (59.2%; 52.5%), 65 to 74 years (29.4%; 31.2%), and 75 years and over (4.5%; 5.8%). The majority of patients did not receive neoadjuvant therapy (86.2%; 88.8%) and had rectal tumors (45.7%; 45.4%). TNM stages included stage I (20.4%; 20%), stage II (36%; 36.9%), stage III (39.1%; 39.7%), and PCR (4.5%; 3.4%). A small proportion of patients in both arms had microsatellite instability–high/mismatch repair–deficient disease (7.3%; 7.5%). Notable alterations included BRAF V600E mutations (2.7%; 2%), KRAS mutations (47.5%; 51.6%), and NRAS mutations (3.8%; 5.1%).

Wading Into Recurrence Data

Patients in the recurrence population had ages ranging from 18 to 44 years (ctDNA arm, 8.3%; surveillance arm, 16.1%), 45 to 64 years (70.8%; 51.6%), and 65 to 74 years (20.8%; 32.3%). Most did not receive neoadjuvant therapy (87.5%; 87.1%). TNM stages included stage I (8.3%; 6.5%), stage II (33.3%; 25.8%), stage III (58.3%; 61.3%), and PCR (0%; 6.5%). Patients had rectal tumors (58.3%; 38.7%), left-sided tumors (37.5%; 38.7%), and right-sided tumors (4.2%; 22.6%). Most patients had adenocarcinoma (91.7%; 89.7%), with a small proportion of patients presenting with mucinosis adenocarcinoma (4.2%; 10.3%) and signet ring cell carcinoma (4.2%; 0%). BRAF/RAS mutations included BRAF V600E (4.2%; 3.6%), KRAS (58.3%; 50%), and NRAS (8.3%; 3.6%).

The median TTCR was 9 months (interquartile range [IQR]), 4.5-11) in the ctDNA arm vs 12 months (IQR, 11-13) in the control arm (lead time, 3.05 months; 95% CI, 1.43-4.67; P < .001). In the ctDNA group, the lead time rates at no more than 2 months, no more than 4 months, and less than 8 months were 24%, 52%, and 75%, respectively.

Regarding recurrent lesions, 62.5% of patients in the ctDNA group had liver and/or lung lesions only vs 51.6% of patients in the surveillance arm. Eleven of 15 patients in the ctDNA arm and 5 of 16 patients in the surveillance arm with only liver and/or lung lesions received curative-intent treatment. Locoregional recurrence was reported in 4.2% of patients in the ctDNA arm and 9.7% of patients in the surveillance group. The lone patient with locoregional recurrence in the ctDNA received curative-intent treatment vs 1 of 3 patients in the control arm.

Other metastatic sites were reported in 33.3% of patients whose disease recurred in the ctDNA arm vs 38.7% of patients in the surveillance arm. No patients in the ctDNA arm received curative-intent treatment for other metastatic sites vs 1 of 12 patients in the surveillance arm.

Among patients with liver metastases in the ctDNA arm (n = 9) and surveillance arm (n = 7), most had 3 or fewer lesions (88.9%; 42.9%). Notably, 77.8% of patients in the ctDNA arm had a tumor size of 3 cm or less vs 42.9% of patients in the control arm. The rates of patients with unilobar lesions were 88.9% and 42.9%, respectively.

“Our ctDNA-guided surveillance identified earlier recurrences amendable to curative therapy at higher rates than standard surveillance, predominantly [with] localized lung/liver metastases,” Peng concluded.

Disclosures: Study co-author Chunming Ding, PhD, is the founder for Innovation Biomed and has filed patent applications for the assay used during the trail study. The remaining study authors did not have any financial disclosures.

Reference

Peng J, Ding C, Mo S, et al. Dynamic circulating tumor DNA methylation monitoring guiding postoperative surveillance in non-metastatic colorectal cancer: Interim analysis of FIND trial. Presented at: 2025 ESMO Gastrointestinal Cancers Congress; July 2-5, 2025; Barcelona, Spain. Abstract LBA1.