The Role of ctDNA in the Management of Non-Small-Cell Lung Cancer in the AI and NGS Era

Abstract

Liquid biopsy (LB) involves the analysis of circulating tumour-derived DNA (ctDNA), providing a minimally invasive method for gathering both quantitative and qualitative information. Genomic analysis of ctDNA through next-generation sequencing (NGS) enables comprehensive genetic profiling of tumours, including non-driver alterations that offer prognostic insights. LB can be applied in both early-stage disease settings, for the diagnosis and monitoring of minimal residual disease (MRD), and advanced disease settings, for monitoring treatment response and understanding the mechanisms behind disease progression and tumour heterogeneity. Currently, LB has limited use in clinical practice, primarily due to its significant costs, limited diagnostic yield, and uncertain prognostic role. The application of artificial intelligence (AI) in the medical field is a promising approach to processing extensive information and applying it to individual cases to enhance therapeutic decision-making and refine risk assessment.

Keywords: NGS; NSCLC; artificial intelligence; ctDNA; liquid biopsy.

Figure 1

Overview of liquid biopsy and ctDNA genomic analysis. ctDNA = tumour-derived circulating DNA; cfDNA = circulating free DNA; PCR = polymerase chain reaction; NGS = next-generation sequencing; ddPCR = digital droplet polymerase chain reaction.

Figure 2

A simple graph to explain the AI workflow from multi-omics data to output and applications. Deep learning is a subfield of machine learning, and machine learning is as a subfield of AI. All algorithms and models should be tested and validated before they are used in clinical trials. Oncologists can apply data derived from AI analyses for a large number of applications, as listed in the graph.

Figure 3

Possible applications in multi-omics fields to dissecting lung cancer and all of the relative diagnostic, prognostic, and therapeutic implications. Artificial intelligence (AI) might support radiologists, pathologists, molecular pathologists, and oncologists, but also surgeons and radiotherapists, in the near future. Liquid biopsy is one fascinating scope for using AI. PFS = progression-free survival; OS = overall survival; MRD = minimal residual disease; LC = lung cancer; TKI = tyrosine kinase inhibitor.

Figure 4

Overview of potential applications of liquid biopsy in NSCLC patient management. (a) Diagnosis: Complementary or exclusive use of LB compared to tissue biopsy. (b) Early-stage treatment escalation/de-escalation based on early molecular response of ctDNA. (c) Monitoring minimal residual disease (MRD) for early detection of disease recurrence. (d) Advanced-stage treatment escalation/de-escalation based on ctDNA levels. (e) Detecting specific resistance mutations to guide targeted therapies.