Objective  To investigate the factors associated with colorectal cancer metastasis and provide clinical reference for predicting the risk of developing metastasis to adjust the treatment strategy in a timely manner.

  Methods  Clinical data were collected from 551 patients who were definitively diagnosed with colorectal cancer using preoperative and/or postoperative pathology in Affiliated Hospital of Yanbian University from June 2012 to June 2022. Multi-factor Logistic regression analysis was performed. The subject receiver operating characteristic (ROC) curve was plotted, analyzing transfer-related factors and their predictive value.

  Results  There were statistically significant differences between the metastatic and non-metastatic groups in tumor type, degree of tumor differentiation, tumor primary site, tumor size, CEA, CA199, CA125, lymphocytes, C-reactive protein, PLR, NLR, CAR, and SII (P<0.05). The areas under the ROC curve for CEA, CA125, C-reactive protein, and PLR, combined, for prediction were 0.633, 0.580, 0.579, 0.599, and 0.665, respectively. The results of multi-factor analysis showed that CEA, C-reactive protein, and PLR were independent risk factors for colorectal cancer metastasis. CEA and CA125 were independent risk factors for colorectal cancer metastasis to the liver only. PLR was an independent risk factor for colorectal cancer metastasis to the lymph nodes only (OR>1, P<0.05).

  Conclusions  CEA, PLR, C-reactive protein, and CEA+CA125+C-reactive protein+PLR have predictive value for colorectal cancer metastasis. CEA and CA125 have predictive value for colorectal cancer metastasis to the liver alone. PLR has predictive value for colorectal cancer metastasis to the lymph nodes alone.