Abstract
BACKGROUND: There is no recognized biomarker is recommended to monitor or predict the prognosis of colorectal cancer (CRC) patients with negative detection of carcinoembryonic antigen (CEA) or carbohydrate antigen 19-9 (CA19-9) and to classify high recurrence-risk cases. METHODS: Discovery and two-stage validation cohorts, which included 2111 radically resected patients with stage II-III CRC, were enrolled in this study. We detected preoperative peripheral monocyte, platelet, albumin (Alb), pre-albumin (pAlb), CEA, and CA19-9 and investigated the prognostic and risk-stratified roles of twelve new inflammatory biomarkers in the three cohorts. RESULTS: In our study, monocyte-to-pAlb ratio (MPAR), monocyte-to-lymphocyte -to-Alb ratio (MLAR), monocyte-to-lymphocyte-to-pAlb ratio (MLPAR), monocyte- to-pAlb score (MPAS), lymphocyte-to-monocyte-Alb score (MLAS), lymphocyte-to monocyte-pAlb score (MLPAS), and platelet-to-lymphocyte-Alb score (PLAS) were significantly associated with both RFS and OS in three cohorts. MLPAS showed the best performance in predicting RFS and OS, and it was related to right-tumor location and significant cancer burden (≥5cm) in the overall population. Moreover, MLPAS is a robust prognostic biomarker in subgroups stratified by CEA or CA19-9. Patients with scores zero and two of the CEA-CA19-9-MLPAS score (CCMLP) showed the lowest and highest recurrence and death rates, respectively, and significant survival differences were observed between them. CONCLUSION: MLPAS is an optimal, independent, and robust prognostic biomarker in the stage II-III CRC population, especially with negative CEA or CA19-9. The CCMLP could effectively classify high recurrence-risk patients who require more focus, monitoring, and treatment for the clinic.