ctDNA as a Prognostic Biomarker in Metastatic Gastric Cancer
High pretreatment ctDNA burden (measured as maximum variant allele frequency or MSAF) independently predicts worse overall survival in metastatic gastric cancer patients receiving first-line chemotherapy-based treatment, and should be used to stratify prognosis at baseline.
Baseline ctDNA Burden: The Strongest Prognostic Signal
The most recent and highest-quality evidence demonstrates that baseline ctDNA burden is an independent prognostic biomarker in metastatic gastric cancer 1. In a 2026 study of 108 patients, those with high MSAF (>12.31%) had significantly worse overall survival compared to low MSAF patients (median OS: 10.0 vs 17.6 months, p=0.006), with high MSAF remaining independently associated with worse OS in multivariate analysis (HR 2.14,95% CI: 1.04-4.41, p=0.039) 1.
This finding is corroborated by real-world data from 824 patients, where high pretreatment ctDNA burden was associated with significantly shorter time to next treatment (4.8 vs 7.4 months, p<0.001) and overall survival (13.2 vs 19.1 months, p<0.001) in patients receiving chemotherapy-based treatments 2.
Key Metrics for Prognostication:
- Maximum variant allele frequency (MVAF) or MSAF: The highest variant allele frequency detected in ctDNA
- Cutoff threshold: Use cohort mean (
12%) or median (2.9%) to stratify high vs low burden 2, 1 - Detection rate: Approximately 75-91% of metastatic gastric cancer patients have detectable ctDNA at baseline 2, 3
Number of Mutations: Additional Prognostic Information
Detection of multiple gene mutations at baseline provides independent prognostic value beyond single-gene analysis 3. Patients with multiple mutations in plasma had worse overall survival (HR 2.16,95% CI 1.10-4.28, p=0.027) and progression-free survival (HR 2.71,95% CI 1.28-5.73, p=0.009) 3.
The number of mutations correlates with treatment response: patients in the ineffective treatment group showed significantly higher numbers of gene mutations (p=0.0006) 4.
On-Treatment ctDNA Monitoring: Early Response Assessment
Residual detectable ctDNA after 9 weeks (3 cycles) of first-line systemic treatment strongly predicts inferior outcomes 3:
- Overall survival: HR 4.95 (95% CI 1.53-16.04, p=0.008)
- Progression-free survival: HR 4.08 (95% CI 1.31-12.75, p=0.016)
Longitudinal monitoring reveals that changes in mutation burden during treatment predict therapeutic efficacy: patients showing increases in gene mutations or copy number levels are more likely to have ineffective treatment (p=0.038) 4.
Clinical Application Algorithm
At Baseline (Before First-Line Treatment):
- Obtain ctDNA testing within 60 days before treatment initiation 2
- Calculate MSAF/MVAF from next-generation sequencing panel
- Stratify patients:
- High burden (MSAF >12%): Expect median OS ~10-13 months
- Low burden (MSAF <12%): Expect median OS ~17-19 months
- Count total mutations detected: Multiple mutations indicate worse prognosis
During Treatment (Week 9):
- Repeat ctDNA testing after 3 cycles of chemotherapy
- Assess ctDNA clearance:
- Undetectable ctDNA: Favorable prognosis, continue current regimen
- Persistent/rising ctDNA: Poor prognosis, consider treatment modification
Ongoing Monitoring:
- Serial ctDNA measurements at 9-week intervals can detect early progression before radiographic changes 3, 4
Important Caveats and Limitations
Treatment-specific considerations: The prognostic value of ctDNA burden appears strongest for chemotherapy-based regimens. In the real-world cohort, ctDNA burden did not show significant associations with outcomes in immunotherapy (n=100) or trastuzumab-based (n=99) treatment groups 2. This suggests ctDNA's prognostic utility may be context-dependent.
Technical factors affecting sensitivity:
- Tumor shedding varies by location and biology 5
- Baseline tumor volume correlates with VAF (Pearson's R 0.53, p<0.0001) 3
- Detection platforms matter: NGS panels targeting frequently mutated genes in gastric cancer provide optimal sensitivity 3
Biological confounders: High MSAF tumors more frequently exhibit deficient mismatch repair and high tumor mutational burden 1, suggesting ctDNA burden reflects underlying biological aggressiveness rather than simply tumor volume.
Guideline Context
While the 2025 NCCN Gastric Cancer Guidelines 6 emphasize the importance of biomarker testing for treatment selection in advanced disease, ctDNA is not yet incorporated into formal staging or risk stratification systems. Current staging classifies all metastatic cases as stage IV without reflecting biological heterogeneity 1. The 2025 ELBS consensus 7 provides technical recommendations for ctDNA reporting but does not establish clinical decision thresholds.
The evidence supports using ctDNA burden as a validated prognostic biomarker to counsel patients on expected outcomes and potentially inform treatment intensity decisions, though prospective interventional trials demonstrating outcome improvement from ctDNA-guided strategies in metastatic gastric cancer are still needed 8.