Circulating tumor DNA (ctDNA) has evolved from an experimental biomarker to an increasingly actionable tool, informing treatment decisions throughout colorectal cancer care. Whether guiding adjuvant therapy intensification or de-escalation, refining organ-preservation strategies in rectal cancer, or identifying subgroups likely to benefit from adjunctive therapies such as celecoxib, ctDNA holds substantial promise for personalizing treatment.
During the 2025 ASCO Gastrointestinal Cancers Symposium, three pivotal abstracts underscored the distinct roles of ctDNA. From stage II or III colorectal cancer in the BESPOKE CRC trial, to measurable residual disease (MRD) detection in node-negative rectal cancer, and finally to predictive benefits for celecoxib in stage III colon cancer, these studies highlight how ctDNA continues to refine risk stratification and guide therapeutic decisions.
Stage II or III Colorectal Cancer
In the BESPOKE CRC trial, Purvi K. Shah, MD, MBBS, of Virginia Cancer Institute, and colleagues employed a commercially available ctDNA assay to assess MRD status in 1,166 patients with resected stage II or III colorectal cancer.1 Oncologists received MRD results and adjusted adjuvant management at their discretion.
Purvi K. Shah, MD, MBBS
Although 84% of oncologists reported that the test affirmed an existing plan, 16% stated that ctDNA findings directly altered the intensity of adjuvant therapy—largely driving de-escalation in ctDNA-negative patients and escalation in ctDNA-positive patients. Of note, 13% of patients had metastatic disease diagnosed earlier because of a positive postoperative ctDNA test, prompting timely therapy for stage IV disease.
According to Dr. Shah, the prognostic value of ctDNA stood out. At the MRD time point (2–6 weeks after surgery), ctDNA positivity in stage II and III disease heralded significantly worse 2-year disease-free survival—45% and 35%, respectively—vs 91% and 87% in ctDNA-negative counterparts. During the surveillance period, the hazard of recurrence was 26-fold higher in patients who developed positive ctDNA at any time compared with those remaining ctDNA-negative.
Additionally, adjuvant chemotherapy benefit was confined in MRD-positive patients, resulting in a 50% lower recurrence risk relative to MRD-positive patients who did not receive chemotherapy. In contrast, MRD-negative patients showed no measurable survival benefit from chemotherapy, hinting that a negative ctDNA test may identify those unlikely to benefit from systemic treatment.
Further underscoring its clinical utility, ctDNA positivity accurately detected recurrence across multiple metastatic sites, including historically low-shedding regions such as the lungs and peritoneum. It also enabled a higher rate of metastasis-directed therapy.
Taken together, said Dr. Shah, these results support ctDNA-driven treatment personalization. For many patients, a negative ctDNA test validated a nonintensive adjuvant plan, mitigating unnecessary toxicity, whereas a positive ctDNA test signaled the need for more aggressive therapy or earlier detection of metastasis.
“With over 70% of participants enrolled from community settings, this large real-world study establishes a practical framework for integrating ctDNA into standard colorectal cancer workflows,” Dr. Shah concluded.
Detecting Residual Disease in Node-Negative Rectal Cancer
The second presentation, led by Jonathan M. Loree, MD, MS, of BC Cancer, examined ctDNA dynamics in patients with T1–3, N0, low- to mid-rectal tumors enrolled in the CCTG CO.28 trial.2 This study employed 3 months of neoadjuvant CAPOX (capecitabine and oxaliplatin) or FOLFOX (leucovorin, fluorouracil, and oxaliplatin), followed by transanal local excision for those demonstrating sufficient downstaging. Patients with inadequate response or high-risk pathologic features were advised to undergo total mesorectal excision.
Jonathan M. Loree, MD, MS
Retrospective ctDNA analyses revealed that 45.8% of patients were ctDNA-positive at baseline, but this rate dropped to 8.7% after chemotherapy. Persistent ctDNA positivity prior to local excision was correlated with unfavorable pathology and a recommendation for more radical surgery. Although the 3-year disease-free survival exceeded 90% across the broader trial population—reflecting its relatively early disease stage—ctDNA signaled which patients were likelier to harbor residual malignancy.
Despite the trial’s small size and infrequent recurrence events, these data suggest ctDNA may aid in patient selection for organ preservation, according to Dr. Loree. Individuals who effectively cleared ctDNA generally needed local excision alone, preserving sphincter function. In contrast, patients who remained ctDNA-positive were more likely to require definitive surgery or additional therapies.
Dr. Loree emphasized the need for larger, prospective trials to confirm how best to integrate ctDNA into early rectal cancer management, but the results already highlight the biomarker’s potential to guide “tailored approaches rather than a one-size-fits-all plan.”
Using ctDNA to Predict Celecoxib Benefit in Stage III Colon Cancer
Jonathan A. Nowak, MD, PhD, of Brigham and Women’s Hospital, presented a post hoc analysis of the CALGB/SWOG 80702 trial, focusing on patients with stage III colon cancer who were randomly assigned to receive celecoxib or placebo with adjuvant FOLFOX.3 Although the overall trial had shown no statistically significant advantage to celecoxib in the full study population, ctDNA testing revealed a subset that appeared to benefit from this anti-inflammatory agent.
Jonathan A. Nowak, MD, PhD
Of 940 evaluable patients, 18% were ctDNA-positive before initiating chemotherapy. These individuals faced a worse prognosis if left on placebo, with a 3-year disease-free survival estimate of 22.6%, compared with 41% for those who received celecoxib. In contrast, ctDNA-negative patients saw minimal difference whether they received celecoxib or placebo, indicating ctDNA positivity might identify a subgroup deriving distinct benefit from COX-2 inhibition.
Additional analyses of tumor molecular features known or hypothesized to be independently predictive of celecoxib benefit, including mismatch repair (microsatellite instability) status and PIK3CA status, did not alter the association between celecoxib and improved survival in ctDNA-positive patients. Dr. Nowak acknowledged the post hoc nature of these findings and the lack of a formal interaction test but notes that the data suggest MRD status may inform the use of celecoxib.
“A ctDNA-positive result potentially signals the presence of micrometastatic disease with an immune microenvironment that is responsive to COX-2 blockade, whereas ctDNA-negative patients likely have a low-risk profile that gains little from adding celecoxib,” Dr. Nowak concluded.
If these early findings are confirmed by future prospective research, ctDNA-guided selection of celecoxib could offer another layer of personalization in stage III colon cancer, allowing oncologists to direct add-on treatments to patients with the highest risk of recurrence.
DISCLOSURE: Dr. Shah reported no conflicts of interest. Dr. Loree reported financial relationships with Advanced Accelerator Applications, Amgen, Bayer, Guardant Health, Ipsen, Merck, Personalis, Roche, and Taiho Pharmaceutical. Dr. Nowak reported financial relationships with Leica Biosystems and Natera.
REFERENCES
1. Shah P, et al: 2025 ASCO Gastrointestinal Cancers Symposium. Abstract 15. Presented January 25, 2025.
2. Loree J, et al: 2025 ASCO Gastrointestinal Cancers Symposium. Abstract 20. Presented January 25, 2025.
3. Nowak J, et al: 2025 ASCO Gastrointestinal Cancers Symposium. Abstract LBA14. Presented January 25, 2025.
