Predictors of Metastatic Lymph Node Burden in Colorectal Cancer: A Negative Binomial Regression Analysis

Background: The prognostic value of preoperative inflammatory and nutritional biomarkers in colorectal cancer has been extensively investigated, but their association with metastatic lymph node burden remains insufficiently explored. This study evaluated whether routinely available biomarkers are associated with the number of metastatic lymph nodes in patients undergoing colorectal cancer resection. Methods: We retrospectively analyzed 935 consecutive patients who underwent colorectal cancer surgery between 2020 and 2024. Preoperative hemoglobin, neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), and Prognostic Nutritional Index (PNI) were assessed. Negative binomial regression was used to model metastatic lymph node count, with sensitivity analyses accounting for surgical setting and total lymph node yield. Results: In the fully adjusted primary model, tumor location (p = 0.021), PNI (adjusted IRR 1.024; p = 0.045), mucinous component (adjusted IRR 1.768; p = 0.026), and lymphovascular invasion (adjusted IRR 2.197; p = 0.012) were associated with metastatic lymph node burden, whereas NLR and PLR were not. After additional adjustment for total lymph node yield, the associations with PNI, mucinous component, and tumor location were attenuated, while lymphovascular invasion remained independently associated with greater nodal burden (adjusted IRR 2.376; p = 0.007). Conclusions: Lymphovascular invasion showed the most robust association with metastatic lymph node burden. Although PNI showed a modest association in the primary model, it provided no significant incremental improvement in model fit, while tumor-location associations were sensitive to model specification. These findings require validation in independent cohorts.