Table 3.

Other biomarkers.

Biomarker	Detection Method	CRC diagnosis AUC	Research Findings
GMSM Model	LASSO algorithm selects metabolite features	0.93 (early-stage CRC)	The GMSM model, correlating serum metabolites with colorectal cancer and adenoma, accurately distinguishes patients from healthy individuals. Its performance significantly exceeds that of CEA and FOBT, holding promise as a noninvasive method for colorectal cancer and adenoma detection (130, 131).
Targeted Metabolite Analysis	Orthogonal partial least squares discriminant analysis and ROC analysis	0.969	Metabolomic analysis reveals different enrichment pathways of intestinal metabolites in colorectal cancer and adenoma. Combined analysis effectively distinguishes patient populations. Certain metabolites, such as 9,10-dihydroxy-12-octadecenoic acid and cholesterol esters, can be used to differentiate colorectal cancer from healthy controls. Relevant metabolites correlate with patient survival (132).
Blood cfDNA Detection	Detects tumor cells, organotypic DNA, and donor DNA in tumors, using CRISPR technology	0.96	Engineered bacteria are employed to detect colorectal cancer cells, organotypic DNA, and donor DNA. This approach holds promise in the diagnosis and treatment of colorectal cancer. Clinically occult and radiologically undetectable minimal residual disease (MRD) during surgery is considered a major source of disease recurrence. Compared to the traditional blood marker CEA, preoperative ctDNA exhibits higher sensitivity, allowing more accurate prediction of recurrence risk (133).
Methylation Biomarker MYO1-G	ddPCR and ROC analysis	0.94	The MYO1-G methylation biomarker, detected using ddPCR technology, accurately distinguishes colorectal cancer patients from normal controls. It is also correlated with disease progression and treatment response. Postoperative monitoring reveals a correlation between methylation levels and patient treatment response and survival, providing a new biomarker for colorectal cancer diagnosis and monitoring (134–136).