Nomogram incorporating prognostic immune-inflammatory-nutritional score for survival prediction in pancreatic cancer: a retrospective study

Abstract

BackgroundPrognosis prediction for pancreatic cancer has always been difficult in clinical practice because of its high heterogeneity and mortality. The aim of the study was to assess the value of prognostic immune-inflammatory-nutritional (PIIN) score on overall survival (OS) in postoperative patients with pancreatic cancer and to develop a nomogram incorporating PIIN score.MethodsThis study retrospectively analyzed the clinic pathological data of 155 patients with pancreatic cancer who underwent radical surgery. PIIN score was calculated by measuring the fibrinogen (FIB), neutrophil to lymphocyte ratio (NLR), systemic immune-inflammation index (SII), albumin-bilirubin (ALBI) score, and prognostic nutritional index (PNI). Patients were divided into two groups by PIIN score levels over a threshold of 37.2. Univariate and multivariate analysis were performed using the Cox regression analysis model. The time-dependent receiver operating characteristic (ROC) curve was plotted to compare the prognostic values of the scoring systems. Finally, a nomogram based on PIIN score was constructed and validated.ResultsMultivariate regression analysis showed that PIIN score (hazard ratio (HR) = 2.171, 95% confidence interval (CI) = 1.207–3.906, P = 0.010), lymphovascular invasion (HR = 1.663, 95% CI = 1.081–2.557, P = 0.021), poor tumor grade (HR = 2.577, 95% CI = 1.668–3.982, P < 0.001), bad TNM stage (I vs. II: HR = 1.791, 95% CI = 1.103–2.906, P = 0.018; I vs. III: HR = 4.313, 95% CI = 2.365–7.865, P < 0.001) and without adjuvant chemotherapy (HR = 0.552, 95% CI = 0.368–0.829, P = 0.004) were independent risk factors for OS. The time-dependent ROC curves revealed that PIIN score was better than the other scoring systems in predicting survival prognosis. And last, the nomogram established from independent factors such as PIIN score had good predictive power for OS. The ROC curve results showed that the AUC values for 1, 3 and 5 years were 0.826, 0.798 and 0.846, respectively. The calibration plots showed the superior clinical applicability of the nomogram.ConclusionThe nomogram model based on PIIN score can be utilized as one of the prognosis stratifications as well as postoperative follow-up for the development of individual treatment for pancreatic cancer.