Non-invasive discrimination of multiple myeloma using label-free serum surface-enhanced Raman scattering spectroscopy in combination with multivariate analysis

We report non-invasive discrimination of multiple myeloma (MM) using label-free serum surface-enhanced Raman scattering (SERS) spectroscopy in combination with multivariate analysis. Colloidal silver nano-particles (AgNPs) were used as the SERS substrate. High quality serum SERS spectra were obtained from 53 MM patients and 44 healthy controls (HCs). The SERS spectral differences demon-strated variation of relative concentrations of biomolecules in the serum of MM patients in comparison to HCs. Multivariate analysis methods, including principal component analysis (PCA), linear discriminant analysis (LDA), and support vector machine (SVM), were used to build discrimination models for MM. Leave-one-out cross-validation (LOOCV) was used to evaluate the performances of the models, in terms of accuracy, sensitivity, specificity, and area under the receiver operating characteristic curves (AUC). Using the SVM model, the accuracy for discrimination of MM was achieved as 78.4%, and the corresponding sensitivity, specificity, and AUC values were 0.830, 0.727, and 0.840, respectively. The results show that the serum SERS in combination with multivariate analysis could be a fast, non-invasive, and cost-effective technique for discrimination of MM. (C) 2021 Elsevier B.V. All rights reserved.

Automated summary

This study reports a non-invasive method for discriminating multiple myeloma (MM) using label-free serum surface-enhanced Raman scattering (SERS) spectroscopy in combination with multivariate analysis. The results show that the relative concentrations of biomolecules in the serum of MM patients differ from those of healthy controls. By building a discrimination model using SVM, an accuracy of 78.4% was achieved for MM discrimination.