Abstract
Lung cancer metastasis is a leading cause of cancer-related mortality, necessitating innovative approaches for early prediction and personalized clinical management. A novel strategy is present to predict lung cancer metastasis risk by combining single-cell metabolic profiling of circulating tumor cells (CTCs) with a self-developed CTC sorting and capture platform, enabling high-efficiency, high-viability CTC isolation from blood. Using nanoelectrospray ionization-atmospheric pressure chemical ionization mass spectrometry, single-cell metabolomic profiling on 301 CTCs derived from patients and animal models are performed. 390 unique metabolites are identified and discovered distinct metabolic signatures associated with different metastatic potentials (brain and bone). Based on these metabolic profiles, a classification model that categorizes CTCs into subgroups with distinct metastatic risks are constructed. The model outperformed traditional clinical indicators and total CTC counts, achieving AUCs of 0.74 (brain metastasis) and 0.92 (bone metastasis). Prospective validation confirmed its metabolite-based classification accuracy for one-year metastasis risk prediction. This study highlights the potential of single-cell metabolomics to uncover novel therapeutic targets and prognostic markers, advancing liquid biopsy from quantitative counting to qualitative analysis. The approach represents a significant advancement in precision medicine for lung cancer management, offering a personalized strategy for predicting metastasis risk and guiding clinical treatment.