Why This Study Matters
Colorectal liver metastasis (CRLM) is a leading driver of colorectal cancer mortality, and up to two-thirds of patients recur after curative-intent liver resection. Whether adjuvant chemotherapy (ACT) improves survival after CRLM resection has been debated for over two decades, since trials have shown disease-free survival gains without a consistent overall survival benefit, leaving clinicians without a reliable way to select which patients actually need postsurgical therapy.
For more than two decades, oncologists have lacked a reliable way to decide which patients with resected colorectal liver metastases (CRLM) truly need adjuvant chemotherapy (ACT) after surgery. This updated analysis from the CIRCULATE-Japan GALAXY registry tests whether a blood-based biomarker — circulating tumor DNA (ctDNA)-defined molecular residual disease (MRD) — can identify the patients who actually benefit from postsurgical treatment, rather than exposing everyone to chemotherapy’s toxicity and cost.
Study Design
Retrospective subset analysis of a prospective, nationwide biomarker registry (CIRCULATE-Japan GALAXY study), landmarked at 70 days postsurgery; registered as UMIN000039205 in the Japan Registry of Clinical Trials (UMIN-CTR) — not a ClinicalTrials.gov-registered trial.
Primary endpoints were disease-free survival (DFS) and overall survival (OS), evaluated via Cox proportional hazards models, stratified by ctDNA-defined molecular residual disease (MRD) status measured 2-10 weeks postsurgery. The treatment comparison was postsurgical adjuvant chemotherapy (ACT) vs. observation (no ACT), compared separately within two cohorts defined by prior treatment history.
Note: This is an observational biomarker cohort analysis, not a randomized controlled trial — there is no randomization scheme or investigational dosing regimen to report; ACT reflects standard-of-care chemotherapy administered at physician discretion.
Patient Population
298 patients with surgically resected colorectal liver metastases (CRLM) without extrahepatic disease, enrolled May 2020-July 2024; median age 67 years (range 33-85); 64.4% male. Split into an upfront surgery cohort (n=191) and a neoadjuvant chemotherapy [NAC] cohort (n=107).
R0 (complete) resection rates were numerically higher in the upfront surgery cohort (184/191, 96.3%) than in the NAC cohort (92/107, 86.0%). ACT was administered to 49/191 (25.6%) of the upfront surgery cohort and 36/107 (33.6%) of the NAC cohort.
Primary Endpoint Results
Subgroup Analyses
| # | Finding |
|---|---|
| 1 | Multivariable analysis, upfront surgery cohort: MRD positivity was the strongest independent prognostic factor for DFS (HR, 7.89; 95% CI, 4.33-14.38; P<.001) and OS (HR, 12.03; 95% CI, 5.08-28.46; P<.001) |
| 2 | Multivariable analysis, upfront surgery cohort, other significant factors for DFS: RAS variants (HR, 1.59; 95% CI, 1.00-2.51; P=.048), liver metastasis size >5 cm (HR, 3.75; 95% CI, 1.39-10.12; P=.009), and ACT (HR, 0.32; 95% CI, 0.17-0.60; P<.001); for OS: RAS variants (HR, 2.49; 95% CI, 1.16-5.35; P=.02) |
| 3 | ctDNA-level subgroup, upfront surgery cohort, MRD-positive patients: ACT improved DFS in both low ctDNA (≤1 MTM/mL; HR, 0.06; 95% CI, 0.01-0.41; P=.004) and high ctDNA (>1 MTM/mL; HR, 0.05; 95% CI, 0-0.54; P=.01) subgroups |
| 4 | ctDNA-level subgroup, NAC cohort: no significant association between ACT and DFS was observed in either low or high ctDNA-level subgroups (exact HR/CI not reported in the source text for this specific comparison) |
| 5 | Longitudinal post-treatment surveillance, upfront surgery cohort: patients who remained serially MRD-negative had improved DFS (HR, 11.27; 95% CI, 6.61-19.21; P<.001) and OS (HR, 12.37; 95% CI, 4.22-36.29; P<.001) versus those MRD-positive at any surveillance time point |
Safety Profile
Not specified in source. This is a retrospective biomarker/prognostic cohort analysis of standard-of-care adjuvant chemotherapy versus observation, not a treatment safety trial, and the full text does not report adverse event or toxicity rates by arm.
This study evaluated standard-of-care adjuvant chemotherapy already established in colorectal cancer treatment, administered under existing safety monitoring practices — it was not designed or powered to report comparative toxicity data by treatment arm.
Interpretation and Broader Context
The findings support a risk-adapted, ctDNA-guided approach to postsurgical treatment after upfront surgery for CRLM: reserving adjuvant chemotherapy for patients with detectable molecular residual disease, since MRD-negative patients showed no significant DFS or OS benefit from ACT. In patients who received neoadjuvant chemotherapy before surgery, ACT was not associated with benefit regardless of MRD status, suggesting current adjuvant strategies may be insufficient in this higher-risk, potentially chemoresistant population.
The authors call for prospective validation of ctDNA-guided adjuvant strategies, particularly in patients with CRLM undergoing upfront surgery, and point to ongoing efforts (e.g., the MONSTAR-SCREEN-3 study and a randomized phase 2 trial of intensified regimens in oligometastatic colorectal cancer) to refine ctDNA detection and treatment escalation strategies for MRD-positive disease.