Impact of methodological variability on the prognostic performance of circulating tumor DNA in colorectal cancer: A systematic review and meta-analysis
Circulating tumor DNA (ctDNA) has demonstrated utility for treatment monitoring, risk stratification, and prognostication in colorectal cancer (CRC). However, substantial methodological heterogeneity across studies may influence its reported prognostic performance. Here, we systematically searched four databases up to October 2025. Eligible studies included patients with CRC who underwent mutation-based plasma ctDNA analysis reported as a binary variable (positive/negative) and provided hazard ratios (HRs) with 95% confidence intervals (95%CIs) for recurrence-free survival (RFS), disease-free survival (DFS), progression-free survival (PFS), or overall survival (OS). Pooled HRs were synthesized using random-effects meta-analysis. Thirty-seven studies met the inclusion criteria. ctDNA positivity was consistently associated with inferior outcomes across all survival endpoints, including OS (HR 3.52; 95%CI 2.42–5.13; p < 0.001; I2 = 74.4%), RFS (HR 4.90; 95%CI 3.54–6.79; p < 0.001; I2 = 83.5%), DFS (HR 4.41; 95%CI 2.96–6.55; p < 0.001; I2 = 84.7%), and PFS (HR 2.46; 95%CI 1.88–3.21; p < 0.001; I2 = 39.2%). The prognostic value of ctDNA was greatest in localized (stage I–III) disease, where it most likely reflects minimal residual disease (MRD). Post-treatment ctDNA demonstrated substantially greater prognostic value than baseline ctDNA, supporting its role in MRD assessment. Persistent ctDNA positivity following surgery and adjuvant chemotherapy identified patients at the highest risk of recurrence and death. Methodological factors were associated with differences in reported prognostic performance: tumor-informed assays outperformed tumor-agnostic assays, particularly in localized disease, while NGS-based approaches showed stronger prognostic performance than dPCR or qPCR. In conclusion, ctDNA positivity is strongly associated with adverse survival outcomes across CRC stages and clinical settings, supporting its role as a clinically relevant biomarker. However, methodological variability significantly influences the reported magnitude of this association, underscoring the need for standardized protocols and transparent reporting to facilitate comparison across studies and support reliable clinical implementation of ctDNA-guided management.