Exploration of the application potential of serum multi-biomarker model in colorectal cancer screening

Abstract

Analyzing blood lipid and bile acid profile changes in colorectal cancer (CRC) patients. Evaluating the integrated model's diagnostic significance for CRC. Ninety-one individuals with colorectal cancer (CRC group) and 120 healthy volunteers (HC group) were selected for comparison. Serum levels of total cholesterol (TC), triglycerides (TG), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), and apolipoproteins (Apo) A1, ApoA2, ApoB, ApoC2, and ApoC3 were measured using immunoturbidimetric and colorimetric methods. Additionally, LC-MS/MS was employed to detect fifteen bile acids in the serum, along with six tumor markers: carcinoembryonic antigen (CEA), carbohydrate antigens (CA) 125, CA19-9, CA242, CA50, and CA72-4. Group comparisons utilized independent sample t-tests and Mann-Whitney U tests. A binary logistic regression algorithm was applied to fit the indicators and establish a screening model; the diagnostic accuracy of individual Indicators and the model was analyzed using receiver operating characteristic (ROC) curves. The CRC group showed significantly lower levels in eight serum lipid indicators and eleven bile acids compared to the HC group (P < 0.05). Conversely, serum levels of TG, CA19-9, and CEA were elevated (P < 0.05). Among the measured parameters, ApoA2 stands out for its strong correlation with the presence of CRC, showcasing exceptional screening efficacy with an area under the curve (AUC) of 0.957, a sensitivity of 85.71%, and a specificity of 93.33%. The screening model, integrating ApoA1, ApoA2, lithocholic acid (LCA), and CEA, attained an impressive AUC of 0.995, surpassing the diagnostic accuracy of individual lipids, bile acids, and tumor markers. CRC patients manifest noteworthy alterations in both blood lipids and bile acid profiles. A screening model incorporating ApoA1, ApoA2, LCA, and CEA provides valuable insights for detecting CRC.

Keywords: Bile acid; Colorectal cancer; Lipid; Screening model.

Figure 1

Flow diagram for development of the CRC screening model. HC Healthy control, CRC Colorectal cancer.

Figure 2

Comparison of lipid Indicators levels between two groups. HC Healthy control, CRC Colorectal cancer; *P < 0.05; **P < 0.01; Error bars (TC, HDL-C, LDL-C, ApoA1, ApoA2, ApoB): Mean with standard deviation; Error bars (TG, ApoC2, ApoC3): Median with interquartile range.

Figure 3

Comparison of bile acids levels between two groups. HC Healthy control, CRC Colorectal cancer; *P < 0.05; **P < 0.01; ns (no statistically significant): P > 0.05; Error bars: Median with interquartile range.

Figure 4

Comparison of tumor markers levels between two groups. HC Healthy control, CRC Colorectal cancer; *P < 0.05; **P < 0.01; ns (no statistically significant): P > 0.05; Error bars: Median with interquartile range.

Figure 5

ROC curve of the model and the markers that make up the model in screening CRC.

Figure 6

The performance of the model in screening CRC. HC Healthy control, CRC Colorectal cancer.