Postsurgical circulating tumor DNA (ctDNA)–defined molecular residual disease (MRD) positivity identified patients with resected colorectal liver metastases (CRLM) who derived an overall survival (OS) benefit from adjuvant chemotherapy, while those with MRD negativity experienced favorable long-term OS without a clear benefit, according to updated data from the CIRCULATE-Japan GALAXY study (UMIN000039205).1,2 These findings were published in JAMA Oncology and presented as an oral presentation during the 2026 ESMO Gastrointestinal Cancers Congress.
Among patients in the upfront surgery cohort who were MRD positive (n = 61), adjuvant chemotherapy was associated with improved OS compared with observation (adjusted HR, 0.27; 95% CI, 0.08-0.88; P = .03). The 48-month OS rates were 65.3% (95% CI, 30.8%-85.7%) vs 32.9% (95% CI, 17.4%-49.3%), respectively. Adjuvant chemotherapy was also associated with improved disease-free survival (DFS) in the same group (adjusted HR, 0.07; 95% CI, 0.02-0.24; P < .001). The respective 48-month DFS rates were 37.5% (95% CI, 15.4%-59.8%) vs not reached (NR). In contrast, among MRD-negative patients in the upfront surgery cohort (n = 128), adjuvant chemotherapy was not associated with differences in DFS (adjusted HR, 0.69; 95% CI, 0.32-1.50; P = .36) or OS (adjusted HR, 0.54; 95% CI, 0.10-2.88; P = .47) compared with observation.
"This analysis adds important OS data in resected CRLM, a setting where clinicians have historically had limited tools to determine who is most likely to benefit from ACT," Adham Jurdi, MD, senior medical director of oncology at Natera, stated in an accompanying news release.²
Previously Reported Findings From the CIRCULATE-Japan GALAXY Study
- Postsurgical ctDNA-defined MRD positivity 2 to 10 weeks after surgery was associated with significantly worse DFS and OS in patients with resectable CRC enrolled in the GALAXY study, confirming MRD as a robust independent prognostic factor beyond conventional clinicopathologic variables.
- In patients with stage II to III CRC, adjuvant chemotherapy was associated with a DFS benefit in those with MRD positivity but not in those with MRD negativity, suggesting the potential for ctDNA-guided, risk-adapted treatment selection.
- These findings laid the foundation for the updated CRLM-specific analysis presented at the 2026 ESMO Gastrointestinal Cancers Congress, which extends the prognostic and predictive utility of postsurgical ctDNA to include OS outcomes.
How was the GALAXY study designed?
GALAXY is the prospective, observational arm of CIRCULATE-Japan, a nationwide registry established to longitudinally assess ctDNA status in patients with clinical stage II to IV colorectal cancer (CRC) undergoing curative resection.1,2 Blood samples were collected at 4, 12, 24, 36, 48, 72, and 96 weeks after surgery until recurrence.
The updated retrospective subset analysis included 298 patients with surgically resected CRLM without extrahepatic disease who had available MRD status 2 to 10 weeks after curative-intent hepatectomy, enrolled between May 2020 and July 2024. Patients with extrahepatic metastases and those enrolled in the interventional trials within the CIRCULATE-Japan platform were excluded. Patients were classified into 2 subcohorts based on whether they had received upfront surgery (n = 191) or neoadjuvant chemotherapy (n = 107).
In this analysis population, the median age was 67 years (range, 33-85), and 192 (64.4%) were male. The median follow-up was 43.2 months (range, 1-68). Adjuvant chemotherapy was administered to 49 patients (25.6%) in the upfront surgery cohort and 36 patients (33.6%) in the neoadjuvant chemotherapy cohort.
MRD status was assessed using Signatera (Natera), a clinically validated, personalized, tumor-informed multiplex polymerase chain reaction next-generation sequencing assay. MRD positivity was defined as detection of at least 2 tumor-specific variants above a predefined threshold. The study’s primary end points were DFS and OS, evaluated using Cox proportional hazards models. Landmark analyses were performed at 70 days postsurgery to account for immortal time bias.
What did the prognostic analyses show?
In the upfront surgery cohort, 62 patients (32.4%) were MRD positive in the 2-to-10-week postsurgical window.1 Compared with MRD-negative patients, those with MRD positivity experienced worse outcomes:
- DFS: HR, 4.14 (95% CI, 2.69-6.36; P < .001); 48-month DFS: 15.6% (95% CI, 6.8%-27.5%) vs 59.6% (95% CI, 50.3%-67.6%); median DFS: 3.8 months (95% CI, 2.8-6.5) vs NR
- OS: HR, 9.13 (95% CI, 4.44-18.78; P < .001); 48-month OS: 41.4% (95% CI, 26.6%-55.7%) vs 91.1% (95% CI, 83.4%-95.3%); median OS: 42.6 months (95% CI, 28.7 to NR) vs NR
In multivariable analysis, MRD positivity was the strongest independent prognostic factor for both inferior 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). Other significant adverse factors included RAS variants for both DFS (HR, 1.59; 95% CI, 1.00-2.51; P = .048) and OS (HR, 2.49; 95% CI, 1.16-5.35; P = .02) and liver metastasis size greater than 5 cm for DFS (HR, 3.75; 95% CI, 1.39-10.12; P = .009).
In longitudinal post–definitive treatment surveillance, patients who remained serially MRD negative had significantly 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) compared with those who were MRD positive at any surveillance time point.
In the neoadjuvant chemotherapy cohort, 51 of 107 patients (47.7%) were MRD positive 2 to 10 weeks after surgery — a higher rate than in the upfront surgery cohort. MRD positivity in the neoadjuvant chemotherapy cohort was similarly prognostic:
- DFS: HR, 4.82 (95% CI, 2.92-7.97; P < .001); 48-month DFS: not reached vs 37.1% (95% CI, 23.9%-50.3%); median DFS: 3.5 months (95% CI, 2.3-6.3) vs 20.7 months (95% CI, 9.7-54.2)
- OS: HR, 9.43 (95% CI, 2.78-31.96; P < .001); 48-month OS: 52.9% (95% CI, 35.1%-67.9%) vs 92.9% (95% CI, 79.4%-97.7%)
Presurgical ctDNA positivity after neoadjuvant chemotherapy completion was also associated with worse OS (HR, 11.42; 95% CI, 1.58-1452.50; P = .009) but not DFS (HR, 1.70; 95% CI, 0.94-3.08; P = .08). Of 18 patients with ctDNA negativity post–neoadjuvant chemotherapy and before surgery, 17 (94%) remained MRD negative at the postsurgical landmark time point.
What did the predictive analysis of adjuvant chemotherapy benefit show?
In the neoadjuvant chemotherapy cohort, adjuvant chemotherapy was not associated with improved DFS or OS regardless of MRD status. Among MRD-positive patients in the neoadjuvant chemotherapy cohort (n = 51), adjuvant chemotherapy did not improve DFS (HR, 0.69; 95% CI, 0.32-1.49; P = .35) or OS (HR, 0.81; 95% CI, 0.27-2.41; P = .70) compared with observation. Among MRD-negative patients in the neoadjuvant chemotherapy cohort (n = 56), outcomes were similarly comparable between adjuvant chemotherapy and observation groups (DFS: HR, 0.63; 95% CI, 0.29-1.37; P = .24; OS: HR, 0.73; 95% CI, 0.07-8.20; P = .80).
Investigators noted that the higher MRD positivity rate in the neoadjuvant chemotherapy cohort may reflect selection of biologically aggressive and chemotherapy-resistant disease and that prior systemic therapy may reduce the incremental benefit of postsurgical adjuvant chemotherapy. An exploratory analysis of ctDNA levels among MRD-positive patients in the upfront surgery cohort found that adjuvant chemotherapy was associated with improved DFS in both low (≤1 mean tumor molecules/mL; HR, 0.06; 95% CI, 0.01-0.41; P = .004) and high (>1 MTM/mL; HR, 0.05; 95% CI, 0-0.54; P = .01) ctDNA level subgroups, suggesting that the presence of detectable ctDNA—rather than absolute ctDNA quantity—captures the primary predictive signal in this setting.
Recurrence patterns also differed by MRD status. Liver metastases at recurrence were more frequent in patients with MRD positivity in both the upfront surgery cohort (70% vs 36%) and neoadjuvant chemotherapy cohort (68% vs 48%), while lung metastases were more common in those with MRD negativity (upfront surgery cohort: 50% vs 13%; neoadjuvant chemotherapy cohort: 38% vs 14%).
What are the implications for practice and next steps?
The GALAXY findings are consistent with emerging prospective data from the phase 2 PKUCRLM-01 study (NCT06404593), in which adjuvant chemotherapy was associated with improved recurrence-free survival and OS in MRD-positive patients with CRLM undergoing upfront surgery, with no benefit observed in those with MRD negativity.1 These data, together with negative or inconclusive ctDNA-guided escalation results from the randomized phase 2/3 DYNAMIC-III trial (ACTRN12617001566325) and the phase 2 FANTASTIC study (jRCTs051190097), suggest that intensification of cytotoxic chemotherapy alone may be insufficient in patients with persistent MRD and that biologically informed strategies—including incorporation of targeted therapies—warrant further evaluation. Prospective validation of ctDNA-guided adjuvant strategies in patients with CRLM, particularly those undergoing upfront surgery, was identified as the priority next step.
References
- Kataoka K, Ito K, Nakamura Y, et al. Circulating tumor DNA status and adjuvant chemotherapy in resected colorectal liver metastases. JAMA Oncol. Published online July 3, 2026. doi:10.1001/jamaoncol.2026.2191
- Signatera™ MRD test predicted overall survival benefit from chemotherapy in resected metastatic colorectal cancer. News release. Natera, Inc. July 6, 2026. Accessed July 7, 2026. https://www.natera.com/company/news/signatera-mrd-test-predicted-overall-survival-benefit-from-chemotherapy-in-resected-metastatic-colorectal-cancer/
