Abstract
Substantial uncertainties remain regarding the establishment of an optimal examined lymph node (ELN) count for the comprehensive management of resectable colorectal cancer (CRC). The correlation of the ELN count with cancer-specific survival (CSS) and benefit from adjuvant chemotherapy (AC) in CRC was investigated using a large database, and the minimal threshold for ELN count in LN-negative patients was determined. The data on stage I to III CRC available in the SEER database (2010-2015) were analyzed to determine the correlation of ELN count with CSS and the benefit of AC using multivariable models. The series of odds ratios (ORs) and hazard ratios (HRs) were fitted using the join-point regression analysis. External validation was performed using the data of patients with stage I to III CRC (2004 to 2009) available in the SEER database. Among LN-negative patients, both cohorts indicated that an increase in the ELN count (≤ 18) led to incremental enhancements in CSS, while no additional improvement in CSS was noted beyond an ELN count of 18. Notably, the efficacy of AC diminished gradually as the ELN count increased. Moreover, post-AC CSS was impaired when the ELN count exceeded 18 (serial HRs > 1), a trend that was accentuated with a higher ELN count. Among the LN-positive patients, two cohorts exhibited proportional increases in ELNs, from one positive LN (PLN) to 20 PLN disease, and incremental benefit from AC, as the ELN count increased (serial HRs < 1). The present study recommends an ELN threshold of 18 when evaluating the quality of prognostic stratification and guiding AC for LN-negative cases. A higher ELN count would be associated with further accurate PLN detection and incremental benefit from AC in LN-positive diseases.