SERS liquid biopsy in colorectal cancer detection and treatment response: Revealing metabolic memory post-radiochemotherapy

Abstract

This study investigates the potential of using surface-enhanced Raman scattering (SERS) serum analysis to assess the response of rectal adenocarcinoma (READ) patients to preoperative radiochemotherapy (RCT). A univariate classification approach differentiated RCT responders (R) from non-responders (NR) with 73 % accuracy. In addition, a classifier trained to differentiate colon cancer from healthy controls was independently applied to the R and NR groups. Using this model, Random Forest identified 86 % of NR samples as cancerous, aligning closely with histopathological findings. Notably, the SERS metabolic profile of the majority of the R sample more closely resembled that of cancer pathology than of healthy controls, suggesting the presence of residual cancer-related metabolic activity, despite the diagnosis of near complete tumor regressions based on histopathology. This user independent classification approach underscores the potential of SERS-based clinical spectroscopy as a non-invasive support tool for predicting tumor response in colorectal cancer.