Predictive ability of prognostic nutritional index in surgically resected gastrointestinal stromal tumors: a propensity score matching analysis

BACKGROUND: Recent findings have shown that inflammation indices are associated with prognosis in various malignancies. However, the usefulness of inflammation indices including platelet-to-lymphocyte ratio, neutrophil-to-lymphocyte ratio and prognostic nutritional index for gastrointestinal stromal tumors (GISTs) remains controversial. METHODS: We retrospectively reviewed 340 primary localized GIST patients who had received surgical resection between 2005 and 2015 to investigate the effect of platelet-to-lymphocyte ratio, neutrophil-to-lymphocyte ratio and prognostic nutritional index on prognosis. 206 patients were selected by propensity score matching to control selection biases. RESULTS: Kaplan-Meier analysis and the log rank test demonstrated that high prognostic nutritional index (≥43.9) was significantly correlated with better recurrence-free survival (RFS) (P<0.001). Among the three inflammatory indices, only preoperative high prognostic nutritional index was an independent prognostic factor for survival [hazard ratio (HR) 0.509; 95% confidence interval (CI) 0.266-0.872; P = 0.031] in multivariate analysis. After propensity score matching, elevated prognostic nutritional index was still a predictor for RFS (HR = 0.498; 95% CI 0.253-0.971; P = 0.042) in the multivariate analyses. In addition, prognostic nutritional index was a significant prognostic factor for GISTs within the National Institutes of Health (NIH) high and very low/low-risk categories. Incorporation prognostic nutritional index into the NIH risk criteria improved the prognostic stratification (c-index, 0.725 vs. 0.686, p = 0.039). CONCLUSIONS: High prognostic nutritional index is a predictor of improved survival for surgically resected GISTs and incorporation prognostic nutritional index into NIH risk criteria improves the predictive accuracy.