Gold nanoparticles based surface-enhanced Raman spectroscopy with multivariate analysis for characterization of 100 kDa blood serum fractions of pancreatic cancer

Abstract

Pancreatic cancer is one of the most aggressive cancers and is often detected at an advanced stage, resulting in poor clinical prognosis. In this study, 100 kDa centrifugal ultrafiltration devices were used to separate filtrate and residue portions from whole serum samples of healthy and pancreatic cancer patients. The filtrate portion, which is high in low molecular weight fractions (LMWF), was analyzed using Surface-enhanced Raman spectroscopy (SERS). This technique utilizes gold nanoparticles as a substrate to analyze filtered fractions. Different spectral differences were seen between cancerous and healthy samples. Principal Component Analysis (PCA) and Partial Least Squares Discriminant Analysis (PLS-DA) were used to classify the SERS spectral data of 100 kDa filtrates of blood serum samples of healthy and different stages of pancreatic cancer. PCA converted the datasets to the most relevant variance by reducing dimensionality. The PLS-DA analysis discriminated between the SERS spectral data into different classes on the basis of their characteristic SERS spectral features. The PLS-DA classification model demonstrated excellent diagnostic performance with 92% sensitivity and 98% specificity and an AUC value of 0.81. These results showed that SERS analysis of ultra-filtered blood serum samples of pancreatic cancer patients is a quick, safe, and accurate way to find biomarkers of this cancer, which could be very useful for early screening of the disease.