Letter to the Editor: Ultra-early adjuvant chemotherapy after gastrectomy: Dose intensity, peritoneal recurrence, and evidence gaps

Table 1 Methodological limitations of current evidence: Sources of bias, clinical implications, and mitigation strategies

Source of bias	Description	Clinical implications	Mitigation strategies
Severe underpowering	Very small ultra-early cohort; wide confidence intervals; post hoc power extremely low in the index study	High risk of false-negative survival conclusions; “no difference” may reflect limited power	Prespecified sample-size calculation; multicenter recruitment; adaptive designs with interim reassessment
Selection bias	Only a small subset receives ultra-early AC; clinician-driven selection; criteria often not explicit	Overestimation of feasibility/safety; limited external validity	Explicit eligibility criteria; screening logs; randomization; intent-to-treat analyses
Residual confounding	Limited matching variables; imbalance may remain in regimen, BMI, histology	Outcomes may reflect baseline differences rather than timing	Randomization; comprehensive baseline capture; multivariable adjustment; sensitivity analyses
Absent biomarker stratification	No MSI/EBV/Lauren or immune/ctDNA assessment	True subgroup effects may be diluted; precision timing not feasible	Mandatory molecular profiling; ctDNA-based MRD stratification; prespecified subgroup analyses
Dose-intensity confounding	More frequent dose reduction; lower median RDI; unclear proactive vs toxicity-driven modifications	Timing effect cannot be separated from dose effect	Prospective dose-finding; standardized dose rules; RDI captured as endpoint with prespecified targets

MSI: Microsatellite instability; EBV: Epstein-Barr virus; BMI: Body mass index; ctDNA: Circulating tumor DNA; RDI: Relative dose intensity.