Abstract
BACKGROUND: Predicting metachronous liver metastasis (MLM) in colorectal cancer (CRC) is crucial in clinical practice. Studies have shown that prealbumin is linked to recurrence and poor overall survival in various malignancies. This study aims to determine whether prealbumin levels can predict postoperative MLM in CRC patients. METHODS: This study included both a primary cohort and an external validation cohort, recruiting a total of 627 CRC patients from two independent hospitals in China. Clinical characteristics and laboratory measurements were collected prior to surgery. The optimal cut-off value for prealbumin in predicting MLM was determined using a receiver operating characteristic (ROC) curve analysis. The Kaplan-Meier method was employed to estimate the 36-month incidence of MLM. The association between prealbumin levels and the occurrence of MLM was evaluated using univariate and multivariate Cox proportional hazard analyses. RESULTS: Postoperative MLM developed in 85 of the 627 patients who underwent surgery. Those who developed MLM had lower prealbumin levels compared to patients without MLM. Kaplan-Meier analysis indicated that patients with reduced prealbumin levels had a higher likelihood of developing MLM. Multivariate Cox regression analysis in both the primary and external validation cohorts confirmed that prealbumin is an independent predictor of MLM. CONCLUSIONS: Prealbumin is an independent predictor for MLM in patients with CRC.