OBJECTIVES: Postoperative delirium (POD) and postoperative cognitive dysfunction (POCD) are common operative neurocognitive disorders, which places a heavy burden on patients, families and society. Therefore, it is very important to search for preventive drugs. Previous studies have demonstrated that perioperative use of dexmedetomidine resulted in a decrease the incidence of POD and POCD. But the specific effect of dexmedetomidine on elderly patients undergoing hepatic lobectomy and its potential mechanism are not clear. This study aims to evaluate the efficacy of intraoperative use of dexmedetomidine on preventing POD and POCD in elderly patients undergoing hepatic lobectomy and the influence on the balance between proinflammation and anti-inflammation. METHODS: This trial was designed as a single-center, prospective, randomized, controlled study. One hundred and twenty hospitalized patients from January 2019 to December 2020, aged 60-80 years old with American Society of Anesthesiologists (ASA) II-III and scheduled for hepatic lobectomy, were randomly allocated into 3 groups (=40) using a random number table: A C group, a Dex1 group, and a Dex2 group. After anesthesia induction, saline in the C group, dexmedetomidine [0.3 μg/(kg·h)] in the Dex1 group, and dexmedetomidine [0.6 μg/(kg·h)] in the Dex2 group were infused until the end of operation. The incidences of hypotension and bradycardia were compared among the 3 groups. Confusion Assessment Method (CAM) for assessing POD and Mini Mental State Examination (MMSE) for evaluating POCD were recorded and venous blood samples were obtained for the determination of neuron specific enolase (NSE), TNF-α, IL-1β, and IL-10 at the different time below: the time before anesthesia (T0), and the first day (T1), the third day (T2), the fifth day (T3), and the seventh day (T4) after operation. RESULTS: Compared with the C group, the incidences of bradycardia in the Dex1 group or the Dex2 group increased (both <0.05) and there was no difference in hypotension in the Dex1 group or the Dex2 group (both >0.05). The incidences of POD in the C group, the Dex1 group, and the Dex2 group were 22.5%, 5.0%, and 7.5%, respectively. The incidences of POD in the Dex1 group or the Dex2 group declined significantly as compared to the C group (both <0.05). However, there is no difference in the incidence of POD between the Dex1 group and the Dex2 group (>0.05). The incidences of POCD in the C group, the Dex1 group, and the Dex2 group were 30.0%, 12.5%, and 10.0%, respectively. The incidences of POCD in the Dex1 group and the Dex2 group declined significantly as compared to the C group (both <0.05). And no obvious difference was seen in the incidence of POCD in the Dex1 group and the Dex2 group (>0.05). Compared with the C group, the level of TNF-α and IL-1β decreased and the level of IL-10 increased at each time points (from T1 to T4) in the Dex1 group and the Dex2 group (all <0.05). Compared with the Dex1 group, the level of IL-1β at T2 and IL-10 from T1 to T3 elevated in the Dex2 group (all <0.05). Compared with the T0, the concentrations of NSE in C group at each time points (from T1 to T4) and in the Dex1 group and the Dex2 group from T1 to T3 increased (all <0.05). Compared with the C group, the level of NSE decreased from T1 to T4 in the Dex1 group and the Dex2 group (all <0.05). CONCLUSIONS: Intraoperative dexmedetomidine infusion can reduce the incidence of POCD and POD in elderly patients undergoing hepatic lobectomy, and the protective mechanism appears to involve the down-regulation of TNF-α and IL-1β and upregulation of IL-10 expression, which lead to rebalance between proinflammation and anti-inflammation.