Molecular-Based Risk Prediction Models for Recurrence After Curative Treatment of Early-Stage Lung Cancer: A...

Recurrence after curative treatment remains a major challenge in early-stage non-small cell lung cancer (NSCLC). Molecular-based prediction models may improve risk stratification and support personalized surveillance strategies. To identify and evaluate molecular-based risk prediction models for recurrence and survival following curative treatment of early-stage NSCLC. A systematic review and meta-analysis were conducted in accordance with PRISMA guidelines and a published PROSPERO protocol. MEDLINE, EMBASE, and the Cochrane Library were searched for studies published between January 2000 and December 2023. Eligible studies reported performance metrics of molecular-based models predicting recurrence-free survival, cancer-specific survival, or overall survival following curative treatment for NSCLC. Data extraction was performed using the CHARMS framework, risk of bias was assessed using PROBAST, and model performance metrics were pooled using random-effects meta-analysis. Of 2447 records identified, five studies met the inclusion criteria. All models used Cox proportional hazards regression. Molecular predictors included mRNA expression profiles (n = 3), long noncoding RNAs (n = 1), and DNA methylation biomarkers (n = 1). Internal validation studies reported AUC values ranging from 0.66 to 0.89, with a pooled AUC of 0.77 (95% CI = 0.66-0.90; I² = 88.4%). Two externally validated models reported AUC values of 0.68-0.74, with a pooled AUC of 0.72 (95% CI = 0.68-0.75; I² = 0%). Two studies were considered high risk of bias due to inappropriate handling of missing data. Molecular-based prediction models demonstrate moderate-to-good discriminative performance for recurrence and survival in early-stage NSCLC. However, broader external validation and improved methodological rigor are required before routine clinical implementation. Trial Registration: PROSPERO identifier: CRD42024599641.