Post-operative ctDNA-detected molecular residual disease in resected non-small cell lung cancer: A systematic review and meta-analysis of prognostic accuracy.

Abstract

e20043 Background: Non-small cell lung cancer (NSCLC) continues to be the most common cause of cancer-related deaths globally, contributing to 85% of all lung cancer cases. Recurrence rates remain high despite complete surgical resection in patients with stage I to III NSCLC, despite appropriate postoperative therapy, occurring in 30-55% of patients. Thus, there is an urgent need for biomarkers like ctDNA to detect molecular residual disease (MRD). In light of the increasing but heterogeneous literature, there is a need for a comprehensive analysis of the prognostic performance of postoperative ctDNA-detected MRD in resected NSCLC. Methods: We searched PubMed, Embase, Scopus, ScienceDirect, and Cochrane from inception to January 2026, identifying 438 records. Eligible RCTs and observational studies reported postoperative ctDNA prognostic accuracy (sensitivity, specificity, NPV, PPV, AUC) and DFS HRs (ctDNA+ vs ctDNA-) in resected NSCLC. Data were pooled using RevMan 5.4 with random-effects models (P<0.05). Results: Across 38 pooled studies (n=4794 resected stage I–III NSCLC), postoperative ctDNA positivity was strongly predictive of recurrence and inferior survival outcomes. The pooled recurrence rate among ctDNA-positive patients was 72.4%, compared with 18.7% among ctDNA-negative individuals. Meta-analysis demonstrated a pooled sensitivity of 0.76 (95% CI, 0.68–0.83), specificity of 0.88 (95% CI, 0.81–0.93), and AUC of 0.90 for predicting recurrence. Postoperative ctDNA positivity was associated with significantly shorter disease-free survival (HR = 3.45; 95% CI, 2.51–4.72; p < 0.001) and overall survival (HR = 2.82; 95% CI, 1.95–4.07). Pooled 2-year DFS rates were 38.1% for ctDNA-positive versus 82.6% for ctDNA-negative patients. Subgroup analyses across stage IB–IIIA disease, adjuvant-treated versus untreated cohorts, and assay type (tumor-informed vs tumor-naïve) demonstrated consistent hazard estimates, with no significant heterogeneity (I² < 40%). Collectively, postoperative ctDNA detection identifies molecular residual disease with high prognostic accuracy and serves as a robust biomarker for early recurrence risk stratification after curative-intent resection. Conclusions: Postoperative ctDNA serves as a reliable biomarker for molecular residual disease in resected NSCLC. ctDNA positivity predicts early recurrence whereas ctDNA negativity predicts durable remission, supporting postoperative surveillance.