Raman consensus estimation for metastatic tumor origin assessment: a proof-of-concept study in colorectal liver metastases.

Determining the tissue of origin of metastatic tumors is essential for tumor classification, origin-directed therapy, and prognostic assessment, particularly in cancers of unknown primary (CUP). Current origin-assessment workflows can be constrained by limited tissue availability, extended turnaround times, infrastructure requirements, and sampling bias. Here, we developed Raman consensus estimation for metastatic tumor origin, a label-free framework that integrates Raman spectroscopic fingerprints with consensus sampling to infer the likely primary origin of metastatic tumors. In a proof-of-concept cohort, the framework correctly identified the primary origin of all clinically confirmed colorectal liver metastases. Consensus voting indicated the feasibility of reduced spectral sampling, suggesting compatibility with clinically relevant time constraints. These findings support Raman consensus estimation as a rapid, non-destructive strategy for metastatic tumor origin assessment, with potential utility in CUP workup, perioperative decision-making, and origin-directed treatment selection.