OBJECTIVES: Circulating tumor DNA (ctDNA) testing has become a promising tool to guide first-line (1L) targeted treatment for advanced non-small cell lung cancer (aNSCLC). This study aims to estimate the clinical validity (CV) and clinical utility (CU) of ctDNA-based next-generation sequencing (NGS) for clinically relevant oncogenic driver mutations to inform 1L treatment decisions in aNSCLC with a systematic literature review and meta-analysis of currently available evidence.

METHODS: A systematic literature search was conducted in PubMed/MEDLINE and Embase on publications between January 2012 and April 2022, to identify randomized control trials or observational studies evaluating CV or CU of ctDNA testing in patients with aNSCLC. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines were followed. Bivariate random-effects models were used for meta-analyses. Pooled estimates on sensitivity and specificity, receiver operating characteristics curves and area under the curve, and progression-free/overall survival (PFS/OS) were reported.

RESULTS: Eighteen studies were identified: 16 CV only, 1 CU only and 1 both. Twelve studies were included for the meta-analysis on multi-gene detection. Overall sensitivity and specificity for ctDNA detection of any mutation were 0.73 (95% CI, 0.64–0.80) and 0.99 (95% CI, 0.96–0.99) respectively. The sensitivity varied greatly between driver genes, from the lowest 0.24 (95% CI: 0.11–0.46) for ROS1 to the highest 0.82 (95% CI: 0.67–0.91) for KRAS. Two studies compared PFS with ctDNA versus tissue-based testing followed by 1L targeted therapy and found no significant differences. No OS data comparing ctDNA versus tissue-based testing were identified from literature.

CONCLUSIONS: ctDNA testing has an overall acceptable diagnostic accuracy in aNSCLC patients, however, sensitivity varies greatly by driver mutation. There were a limited number of studies on the more uncommon driver mutations (i.e., RET, ROS1, NTRK), and this is an essential area of future research.