Abstract: Objective To construct and validate a predictive model  based on the physiological and surgical severity score (POSSUM) system and inflammatory factors for assessing the risk of postoperative complications in elderly patients with locally advanced esophageal squamous cell carcinoma(ESCC).Methods From January 2021 to September 2024,197 elderly patients with locally advanced ESCC were admitted,among whom 105 cases developed postoperative complications.Patients were divided into non-severe (n=65,grade 1-2) and severe (n=40,grade 3-5) complication groups based on Clavien-Dindo grading.Logistic regression identified independent risk factors for serious complications.ROC and AUC assessed diagnostic efficacy of POSSUM and inflammatory factors.Based on these factors,a nomogram risk prediction model was constructed,and the prediction efficacy,calibration degree and clinical utility of the model were evaluated by AUC,calibration curve and DCA.Results Multivariate analysis showed POSSUM score and inflammatory factors (TNF-α,IL-8) were independent risk factors for serious complications (P<0.05).ROC curve analysis showed that POSSUM score,TNF-α,IL-8 combined prediction of AUC was significantly higher than used alone POSSUM score(Z=2.515,P=0.012),TNF-α(Z=2.725,P=0.006) and IL-8(Z=3.579,AUC of P<0.001(P<0.05)).Compared to POSSUM alone,the joint model had NRI=0.35 (P<0.05) and IDI=0.12 (P<0.05).The nomogram model had good fit,threshold probabilities,and clinical net benefits.Conclusion The nomogram model based on POSSUM and inflammatory factors (TNF-α,IL-8) has high predictive value and clinical utility for postoperative complications in elderly locally advanced ESCC patients.

Key words: POSSUM system, inflammatory factors, elderly locally advanced esophageal squamous cell carcinoma, serious postoperative complications, predictive model, validation