Importance: Serial circulating tumor DNA (ctDNA) has emerged as a routine surveillance strategy for patients with resected colorectal cancer, but how serial ctDNA monitoring is associated with potential curative outcomes has not been formally assessed.
Objective: To examine whether there is a benefit of adding serial ctDNA assays to standard-of-care imaging surveillance for potential curative outcomes in patients with resected colorectal cancer.
Design, setting, and participants: In this single-center (City of Hope Comprehensive Cancer Center, Duarte, California), retrospective, case cohort study, patients with stage II to IV colorectal cancer underwent curative resection and were monitored with serial ctDNA assay and National Cancer Center Network (NCCN)-guided imaging surveillance from September 20, 2019, to April 3, 2024. The median duration of follow-up was 26 months (range, 2-54 months).
Interventions: Serial ctDNA assays were performed every 3 months for 2 years and every 6 months for the 3 following years in conjunction with NCCN-guided radiographic surveillance.
Main outcomes and measures: The primary outcome was the proportion of patients with clinical benefit from ctDNA testing, defined as the proportion of patients with a newly positive ctDNA assay and negative scheduled imaging (most recent or concurrent) that subsequently led to early imaging confirmation of recurrence, followed by curative-intent intervention with no evidence of recurrence at the time of data cutoff. Recurrence was categorized by ctDNA recurrence, radiographic recurrence, or concurrent ctDNA and imaging recurrence. Salvage resections and associated durable remissions were described within each of the 3 categories. Descriptive statistics were used to characterize the patient population.
Results: In total, 184 patients (median age, 59 years [range, 32-88 years]; 97 female [52.7%]) were included in this study, and 129 (70.1%) had stage II to III disease. Forty-five patients (24.5%) had ctDNA or imaging-confirmed recurrence. Of these 45 patients, 14 had radiographic recurrence with negative ctDNA, and 11 had concurrent ctDNA and imaging recurrence. Twenty of 45 patients had ctDNA positivity with negative imaging at first ctDNA positivity; 6 had reflex imaging that was positive for recurrence, and 14 continued with serial imaging and ctDNA monitoring. Ten of 14 patients had subsequent recurrent disease, 3 patients had a spontaneous clearance of ctDNA, and 1 patient remained imaging negative 7 months after positive ctDNA, after which she was lost to follow-up. Altogether, 11 of 20 patients with ctDNA recurrence without initial concurrent imaging recurrence had subsequent metastasectomy, and only 3 were disease-free at the cutoff date in April 2024, representing 1.6% of the surveilled population.
Conclusions and relevance: In this cohort study of patients with stage II to IV colorectal cancer who underwent curative-intent resection, the addition of serial tumor-informed ctDNA assay to the standard NCCN-recommended surveillance had limited clinical benefits. Additional prospective research is needed to clarify the value of ctDNA testing in the surveillance setting.