Abstract: Objective: To explore the risk factors of Postoperative Cognitive Dysfunction (POCD) in patients with primary Hepatocellular Carcinoma (HCC) and develop an early warning model for such patients. Methods: We selected 211 patients with HCC who underwent radical resection of liver cancer from January 2019 to January 2021 in a tertiary grade A hospital in Haikou city. The ROC curve was used to analyze the single factor to observe the difference of continuous variables and obtain the best cut-off value. Multivariate Logistic regression model was used to figure out the risk factors of POCD in HCC patients. Risk factors were used as predictive indicators for constructing the Nomogram model. Internal data was adopted to verify the performance of the Nomogram model and the decision curve analysis was used to evaluate the clinical net benefit of the Nomogram model. Results: Among 211 patients with HCC, 47 patients were diagnosed with POCD (22.3%). Independent risk factors of POCD in HCC patients included age (>66 years old), American Society of Anesthesiologists (ASA) classification (Class III), anesthesia time (>270 min), NLR (>2.68) and TNF-α (>9.87 nmol/L) (all P<0.05). The C-index predicted by the Nomogram model was 0.735 (95% CI: 0.704, 0.847); when the Nomogram model predicted a risk threshold of POCD in HCC patients higher than 0.16, significant additional clinical net benefits would occur. In addition, the clinical net benefit of the Nomogram model was higher than age, ASA classification, anesthesia time, TNF-α and NLR. Conclusion: The early warning model constructed in this study can help predict the risk of POCD in patients undergoing radical HCC surgery and thus provide a method for medical staff to formulate POCD prediction strategies.

Key words: primary hepatocellular carcinoma; radical resection; cognitive dysfunction; Nomogram