Abstract
OBJECTIVES: The purpose of this investigation was to assess the prognostic importance of the combination of preoperative hemoglobin (Hb) levels and Geriatric Nutritional Risk Index (GNRI) in forecasting postoperative survival outcomes for patients undergoing pancreaticoduodenectomy (PD) due to Vater ampulla carcinoma (VPC). METHODS: The GNRI nutritional screening was conducted for all patients, and patient outcomes, including overall survival (OS), were subsequently monitored. An H- GNRI scoring system was established using the optimal critical values of 125.5 g/L for Hb and 91.72 for GNRI, as determined by X-tile software. According to the H-GNRI score, the patients were categorized into three groups, namely low H-GNRI group (H-GNRI score = 0, n = 47) with Hb < 125.5 g/L and GNRI < 91.72; medium H-GNRI group (H-GNRI score = 1, n = 77) with Hb < 125.5 g/L and GNRI ≥ 91.72, or Hb ≥ 125.5 g/L and GNR < 91.72; and high H-GNRI group (H-GNRI score = 2, n = 51) with Hb ≥ 125.5 g/L and GNRI ≥ 91.72. The Kaplan-Meier analysis and log-rank tests were employed to evaluate the OS rate and compare survival disparities among various groups. Additionally, both univariate and multivariate analyses were conducted utilizing the Cox regression model, with p < 0.05 considered statistically significant. Finally, to evaluate the predictive effectiveness of Hb, GNRI, and H-GNRI, a receiver operating characteristic (ROC) curve was constructed to compare the area under curve (AUC) values. RESULTS: The OS rate was higher in patients with high Hb levels (≥ 125.5 g/L) compared to those with low Hb levels (< 125.5 g/L). Likewise, patients in the high GNRI group (≥ 91.72) exhibited significantly superior OS compared to those in the low GNRI group (< 91.72). Compared with both the medium and low H-GNRI groups, the high H-GNRI group demonstrated a notably higher OS rate. The T stage (HR = 2.523, 95% CI: 1.694-3.757, p < 0.001), N stage (HR = 2.018, 95% CI: 1.255-3.246, p = 0.004), and the H-GNRI score (H-GNRI score of 2 used as the baseline; H-GNRI score of 0: HR = 2.569, 95% CI: 1.499-4.402, p < 0.001; H-GNRI score of 1: HR = 1.835, 95% CI: 1.118-3.014, p = 0.016), after adjusting for gender, were determined to be independent significant predictors affecting the OS of patients with VPC. The AUC of H-GNRI was 0.677, exceeding that of Hb levels (0.631) and GNRI (0.615). CONCLUSIONS: The combination of preoperative Hb levels and GNRI demonstrates superior predictive efficacy for VPC patients undergoing PD, compared with either Hb levels or GNRI score alone. Therefore, the H-GNRI score can be utilized to promptly identify high-risk patients, establish comprehensive nutritional pre-rehabilitation plans through interdisciplinary collaboration, and inform decisions regarding additional adjunctive therapies.