Perioperative dynamics of the prognostic nutritional index predict recurrence and survival after resection for colorectal liver metastases

We evaluated how perioperative changes in the prognostic nutritional index (PNI) influence recurrence and survival after hepatectomy for colorectal cancer liver metastases (CRLM). We conducted a retrospective analysis of 108 patients who underwent hepatectomy for CRLM. Preoperative and postoperative PNI levels were compared, and patients were categorized into high (postoperative PNI equal to or higher than preoperative PNI) and low (postoperative PNI lower than preoperative PNI) groups. Statistical analyses included the Kaplan-Meier analysis, log-rank tests, Cox proportional hazards regression models, and Akaike information criterion. The median observation period was 3.3 years. The Kaplan-Meier analysis showed worse overall survival and recurrence-free survival in the low group. Subgroup analysis demonstrated reproducibility regardless of patient background, era of surgery, surgical outcomes, neoadjuvant or adjuvant chemotherapy, and liver metastasis status. Multivariate analysis identified low group assignment and synchronous liver metastasis as independent predictors of recurrence (hazard ratio [HR], 3.788; 95% confidence interval [CI], 2.030-7.621; HR, 3.338; 95% CI, 1.715-6.964, respectively) and survival (HR, 4.484; 95% CI, 1.945-10.334; HR, 2.529; 95% CI, 1.183-5.407, respectively). The Akaike information criterion analysis showed that perioperative shifts in PNI were superior to preoperative or postoperative PNI alone in predicting outcomes. Perioperative shifts in PNI predicted recurrence and prognosis post-hepatectomy for CRLM more accurately compared to using preoperative or postoperative PNI alone.