Prognostic value of circulating tumor DNA in operable non-small cell lung cancer: a systematic review and reconstructed individual patient-data based meta-analysis

BACKGROUND: This reconstructed individual patient data (IPD)-based meta-analysis is aimed to summarize the current findings and comprehensively investigate the predictive value of circulating tumor DNA (ctDNA) in operable non-small cell lung cancer (NSCLC). METHODS: PubMed, Cochrane and Embase were searched to include potentially eligible studies. The primary outcomes included progression-free survival (DFS) by ctDNA status at baseline, postoperative, and longitudinal timepoints. The IPD-based survival data was retracted and used in reconstructed IPD-based meta-analysis. Subgroup analysis was implemented based on the baseline characteristics. RESULTS: Totally, 28 studies were involved, including 15 full-length articles (1686 patients) for IPD-based synthesis and 20 studies for conventional meta-analysis. The IPD-based meta-analysis discovered that patients with positive ctDNA status at the baseline (hazard ratio, HR = 3.73, 95% confidential interval, CI: 2.95–4.72), postoperative (3.96, 2.19–7.16), or longitudinal timepoints (12.33, 8.72–17.43) showed significantly higher risk of recurrence. Patients with persistent ctDNA-negative status had the lowest recurrence rate, and the negative conversion of ctDNA from baseline to postoperative timepoints was correlated with elevated DFS. Subgroup analyses suggested that stage II–III patients with ctDNA-positive status may achieve preferable therapeutic outcomes. CONCLUSIONS: Plasm ctDNA monitoring shows excellent clinical significance at the tested timepoints. Perioperative conversion of ctDNA status may indicate the therapeutic effect of radical surgery. Postoperative adjuvant therapy may be determined according to the ctDNA status. TRAIL REGISTRATION: CRD42022304445. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12916-023-03181-2.