Establishment of an inflammatory cytokine-based predictive model for the onset of prolonged postoperative ileus after radical gastrectomy: a prospective cohort study.

Background: Prolonged postoperative ileus (PPOI) is a common postoperative abdominal complication and is strongly associated with the inflammatory response. However, there is a lack of effective means to predict PPOI in patients with gastric cancer. Methods: 222 patients underwent radical gastrectomy at our center were enrolled and divided into the training group and validation cohort. Receiver operating characteristic (ROC) curve analysis and univariate and multivariable logistic regression models were performed to help filter variables for inclusion in the predictive model. And then a nomogram for PPOI was established. The area under the ROC curve (AUC) was calculated to assess the prediction accuracy. Diagnostic calibration curves were used to assess the goodness-of-fit of the nomogram. Decision Curve Analysis (DCA) was applied to evaluate its clinical utility. Results: Significant increase of IL-6, IL-10, TNF-α, and CRP on the first postoperative day were found in PPOI patients after surgery. Univariate and multivariate analysis demonstrated that age ≥ 65, IL-6, and IL-10 were independent predictive factors for PPOI. We subsequently developed a prediction nomogram of PPOI which included age, IL-6, IL-10, and TNF-α. Further verification by the training and validation groups demonstrated the good predictive efficacy of our model, as well as favorable clinical benefits. Conclusions: We developed a novel and easy-to-use prediction nomogram for gastric cancer, which was primarily based on the postoperative level of inflammatory mediators. This model provided further clarification of the exact relationship between inflammatory factors and the occurrence of PPOI, and help us clinically identify the high-risk groups of PPOI for the purpose of early intervention.