Raman spectroscopy on dried blood plasma allows diagnosis and monitoring of colorectal cancer

Abstract

Colorectal cancer (CRC) remains challenging to diagnose, necessitating the identification of a noninvasive biomarker that can differentiate it from other conditions such as inflammatory bowel diseases (IBD) and diverticular disease (DD). Raman spectroscopy (RS) stands out as a promising technique for monitoring blood biochemical profiles, with the potential to identify distinct signatures identifying CRC subjects. We performed RS analysis on dried plasma from 120 subjects: 32 CRC patients, 37 IBD patients, 20 DD patients, and 31 healthy controls. We also conducted longitudinal studies of CRC patient's postsurgery to monitor the spectral changes over time. We identified six spectral features that showed significant differences between CRC and non-CRC patients, corresponding to tryptophan, tyrosine, phenylalanine, lipids, carotenoids, and disulfide bridges. These features enabled the classification of CRC patients with an accuracy of 87.5%. Moreover, longitudinal analysis revealed that the spectral differences normalized over 6 months after surgery, indicating their association with the presence of the disease. Our study demonstrates the potential of RS to identify specific biomolecular signatures related to CRC. These results suggest that RS could be a novel screening and monitoring tool, providing valuable insights for the development of noninvasive and accurate diagnostic methods for CRC.

Keywords: Raman spectroscopy; biomarkers; colorectal cancer; dried blood plasma; inflammation.

FIGURE 1

(A) Average normalized Raman spectra of all dry plasma samples from the different cohort of subjects included. HCs, light blue; DD, light green; IDB, dark green; CRC, light red. (B) differential spectra between DD and HCs; (C) differential spectra between IDB and HA; (D) differential spectra between DVD and HCs.

FIGURE 2

Box plot showing the different intensities of selected peaks between the different groups. Each data point represents an individual subject analysed. Each box represents the area between the 25th and 75th percentiles. Lines inside the boxes represent the median values. Upward arrows signify an increase, while downward arrows indicate a decrease in the intensities of selected peaks over the specified conditions DD, IBD and CRC compared with the HCs. Statistical significance: *p < 0.05; **p < 0.01; ***p < 0.0005.

FIGURE 3

(A) Multivariate analysis of Raman spectra. Linear projection of the six characteristic features using the Freeviz algorithm. (B) The confusion matrix of the classification as number of patients. (C) The confusion matrix of the classification as proportion of actual patients.

FIGURE 4

Box plots showing the different Raman intensity of selected peaks in CRC patients versus HCs over the time. Data are shown as box and whisker plots. CRC, preoperative patients; 6 days (6d), 30 days (30d), and 6 months (6 m) CRC patients after the surgery treatment. Each box represents the area between the 25th and 75th percentiles. Lines inside the boxes represent the median values. Upward arrows signify an increase, while downward arrows indicate a decrease in the intensities of selected peaks over HCs. Statistical significance: *p < 0.05; **p < 0.01; ***p < 0.0005.