Effects of Dexmedetomidine on Perioperative Serum Potassium and Postoperative Rehabilitation of Patients Undergoing Radical Resection of Gastrointestinal Malignancy

Shunping Tian the A liated Hospital of Yangzhou University, Yangzhou University Dongsheng Zhang the A liated Hospital of Yangzhou University, Yangzhou University Ying Wang Dalian Medical University Leyang Yu Dalian Medical University Yanlong Yu the A liated Hospital of Yangzhou University, Yangzhou University Ning Li the A liated Hospital of Yangzhou University, Yangzhou University Shiting Yan Dalian Medical University Wei Zhou the A liated Hospital of Yangzhou University, Yangzhou University Zhuan Zhang (  zhangzhuancg@163.com ) the A liated Hospital of Yangzhou University, Yangzhou University


Introduction
Dexmedetomidine is a highly selective α 2 -adrenergic receptor agonist, which is often used in clinical anesthesia and intensive care unit due to its sedative, analgesic and sympathetic suppressing effects [1]. For radical gastrointestinal malignancy surgery with long operation time and heavy trauma, dexmedetomidine is often used during general anesthesia to blunt strong stress response. Patients with gastrointestinal malignant tumors often harbor several risk factors such as advanced age, poor nutritional state owing to appetite loss and combined underlying diseases [2], which, with prolonged preoperative fasting, are prone to inducing electrolyte disorders.
While prolonged use of dexmedetomidine is associated with a risk of prolonging QT interval and causing the incidence of tip torsion ventricular tachycardia [3]. Perioperative hypokalemia might occur during gastrointestinal surgery and affect myocardial repolarization to increase the risk of tip torsion ventricular tachycardia [4]. When severe, potassium disorders can lead to life-threatening cardiac conduction disturbances and neuromuscular dysfunction [5]. It is still not clear whether prolonged use of dexmedetomidine during gastrointestinal malignancy radical resection could induce or exacerbate hypokalemia or not.
The primary aim of this prospective, randomized, doubled-blind study is to observe the effects of prolonged use of dexmedetomidine on electrolytes concentration and postoperative rehabilitation. The second aim is to explore the optimal dose of dexmedetomidine with the least effect on serum potassium concentration in patients undergoing radical resection of gastrointestinal tumors.

Materials And Methods
This study was approved by the Ethics Committee of the A liated Hospital of Yangzhou University (ethics number: 2017-YKL11-01) and registered on the clinicaltrials.gov website (registration number: NCT04771637). After obtaining informed written consent for participation in the study, patients aged 55-75 years of age, body mass index (BMI) 18.5-25 kg/m 2 and American Society of Anesthesiologists status I-III of both genders underwent general anesthesia for elective radical resection of gastrointestinal malignancy with an estimated operation time longer than 3 hours were enrolled in this study between January 2017 and October 2020. Patients were not admitted to the study if any of the following criteria were present: severe sinus bradycardia (heart rate < 50 beats/min), pacemaker implantation, atrioventricular block, severe circulatory disturbance during operation, malignant arrhythmia, impairment of liver or kidney, preoperative severe internal environmental disorders such as hypokalemia (serum potassium level < 3.0 mmol/L), hyperkalemia (serum potassium level > 5.5 mmol/L), hyperlactatemia (blood lactate concentration > 2.0 mmol/L), intraoperative blood transfusion and postoperative need to be transferred to intensive care unit for further treatment.

Randomization and blinding
Eligible patients were randomly assigned to one of four groups according to a computer generated table of random numbers: groups D 1 , D 2 and D 3 received dexmedetomidine loading dose 1, 1 and 0.5 μg/kg, which was infused intravenously for 10 min, and maintenance dose 0.25, 0.5 and 0.5 μg/kg/h, respectively, group C received normal saline 50 ml/h for 10 min and maintenance dose 10 ml/h. Dexmedetomidine and normal saline maintenance doses in the four groups were continued until approximately 30 min before surgery completion. The sealed opaque envelope technique was used for group allocation. A researcher not involved in the study implementation opened the envelopes. Patients, surgeons, and the outcome assessor were blinded to the technique used.

Anesthesia
None of the patients received preoperative medication. Routine monitoring including electrocardiograph (ECG), pulse oxygen saturation (SpO 2 ) and noninvasive blood pressure (NBP) were performed after patients entering the operating room. Narcotrend monitoring (MT MonitorTechnik GmbH&Co.KG, Germany) was used to monitor the depth of anesthesia. Ultrasound-guided right internal jugular vein puncture was performed under local anesthesia, and a double-lumen catheter was inserted to a depth of 12-13 cm. After that, a combination of 0.9 % sodium chloride solution and 6% hydroxyethyl starch solution 130/0.4 at a ratio of 1:2 was infused at a rate of 6-7 mL/kg/h. Patients with a normal result on the modi ed Allen's test were catheterized with radial artery puncture under local anesthesia to monitor invasive arterial pressure (IAP). Vital signs [mean blood pressure (MAP) and heart rate (HR)] were recorded 5 min after arteriovenous puncture. Groups D 1 , D 2 , and D 3 were given an intravenous injection of dexmedetomidine (18092931,19012631,19081331, Yangzijiang Pharmaceutical Co., Ltd.) with the loading dose for 10 min before general anesthesia induction and the maintenance dose until approximately 30 min before the end of surgery. Group C was given intravenous normal saline at a rate of 50 mL/h for 10 min, followed by continuous infusion at a rate of 10 mL/h until about 30 min before the end of surgery. Dexmedetomidine or normal saline was prepared by the study personnel who did not participate in data collection, and the syringe pumped through the fourth channel of a syringe pump (CP-2100, Beijing Silugao Medical Technology Co., Ltd., China) was covered with a piece of opaque sheet.
General anesthesia induction was carried out with midazolam 0.05 mg/kg, sufentanil 4 μg/kg and propofol 1.5 mg/kg in all groups. Endotracheal intubation and mechanical ventilation were performed after su cient muscle relaxation with cisatracurium 0.2 mg/kg. Mechanical ventilation was maintained with inhaled oxygen concentration 60%, tidal volume 6-8 mL/kg, ventilation frequency 12-14 times/min, and inspiration/expiration ratio 1:1.5 to maintain P ET CO 2 at 35-45 mmHg. Temperature was maintained at 36°C-37°C. General anesthesia maintenance was implemented with intravenous propofol and remifentanil and inhaled sevo urane 1 MAC in all groups. Muscle relaxation was maintained with cisatracurium infusion at 0.15 mg·kg -1 ·h -1 . The doses of propofol and remifentanil were adjusted according to the Narcotrend monitoring, the stage of which was maintained between D2 to E1 (sedation index 20-46) throughout surgery in all groups. Train of four (TOF) monitoring was used to monitor the muscle relaxation during the surgery. When the changes of IAP or HR were >20% of basic levels, prepared vasoactive drugs including norepinephrine, ephedrine or atropine was used. Sevo urane and cisatracurium were stopped at about 30 minutes before surgery ending, and propofol and remifentanil were continued until the end of surgery. Residual neuromuscular blockade was antagonized at TOF ratio more than 0.75 with intravenous neostigmine 0.04-0.07 mg/kg and atropine 0.01-0.02 mg/kg. Patients were extubated when achieved extubation conditions and transferred to PACU thereafter.

Outcome variables
MAP and HR were recorded at 5 min after artery puncture (T 1 ), dexmedetomidine loading dose nish (T 2 ), surgery beginning (T 3 ), 5 min (T 4 ), 10 min (T 5 ) and 1 h (T 6 ) after surgery beginning, surgery ending (T 7 ), extubation (T 8 ), 5 min after extubation (T 9 ) and 1 hour after into PACU (T 10 ). Serum potassium and lactic acid were recorded at T 1 , T 6 , T 7 and T 10 . Serum potassium was also examined at 24 h postoperatively (T 11 ). Liver and kidney function, intraperitoneal drainage volume at 24 h postoperatively, length of hospital stay and postoperative complications such as pulmonary infection were also recorded.

Statistical analysis
The necessary sample size was calculated using PASS 15.0 (NCSS Llc., USA). The sample size calculation was based on the primary endpoint of this study, serum potassium concentration. Based on our pilot study, serum potassium concentrations in patients treated with dexmedetomidine infusion were found to have an average increase of 0.24 mmol/L with a standard deviation of 0.36 mmol/L. Consequently, enrollment of 25 patients in each group would obtain a power of 80% (β=0.2) at a signi cant level of 0.05 (α=0.05, two-tailed). Considering study withdrawals or protocol violation, recruitment of 30 in each group was determined.
Statistical analysis was done using SPSS 25.0 (SPSS Inc., USA). Parametric data were expressed in means ± standard deviation (SD), and Student's t test was used for analyses. Nominal variables were compared using chi-square test. Changes in mean blood pressure and heart rate among groups were analyzed by repeated measures ANOVA. P<0.05 was considered signi cant.

Page 5/16
A total of 140 patients were assessed for eligibility, and 124 subjects were enrolled in the study. The remaining 16 patients were excluded because either they did not meet the inclusion criteria (n=10) or they refused to participate (n=6). Thirty-one patients were allocated to each group. Due to protocol violation, 1 patient were excluded from group D 2 . Therefore, 31 patients in group D 1 , 30 patients in group D 2 , 31 patients in group D 3 , and 31 patients in group C were analyzed ( Figure 1).
There were no signi cant differences of the demographics among the four groups (Table 1). Compared with group C, urine volume increased signi cantly in groups D 2 and D 3 (P=0.004 and 0.039, respectively), while other intraoperative data were similar among the four groups (Table 2).

Potassium and lactate
Compared with the basic serum potassium values, the concentrations at T 11 increased signi cantly in groups D 1 , D 2 and C (P=0.003, 0.002, and 0.001, respectively), while increased signi cantly at T 7 (P=0.008), T 10 (P=0.015) and T 11 (P 0.001) in group D 3 . Compared with group C, the serum potassium concentration in group D 2 decreased signi cantly at T 11 (P=0.032). Compared with group D 2 , the serum potassium concentration in group D 3 increased signi cantly at T 7 (P=0.036).
Compared with the basic serum lactate values, the concentrations decreased signi cantly at T 6 in groups D 2 and D 3 (P=0.002 and 0.003, respectively). Compared with group D 2 , the serum lactate concentrations in group D 3 decreased signi cantly at T 10 (P=0.033) ( Table 3).

Hemodynamic results
Repeated-measures analysis of variance revealed signi cant variation in HR and MAP over time within each group ( Table 4, 5).

Perioperative drug use
Signi cantly, less dosage of propofol was required (P=0.011) in group D 2 as compared to Group C. Again, dose of remifentanil was signi cantly lower in groups D 1 , D 2 and D 3 (P=0.039, 0.004 and 0.042, respectively) than group C. More patients in group C received ephedrine to treat intraoperative hypotension (P=0.048) than the other groups.

Postoperative recovery indicators
The activity of plasma alanine aminotransferase (ALT) was signi cantly higher at T 11 than T 1 in all groups (P=0.030, 0.001, 0.005 and 0.011 in groups D 1 , D 2 , D 3 and C, respectively), and plasma aspartate aminotransferase (AST) activities increased signi cantly at T 11 than T 1 in groups D 2 , D 3 and C (P=0.037, 0.034 and 0.031, respectively). There was no signi cant difference in serum creatinine. Additionally, patients in group D 1 had signi cantly higher blood urea nitrogen at T 11 than T 1 (P=0.014) ( Table 6). There were no signi cant differences in length of hospital stay, intraperitoneal drainage volume and pulmonary infection incidence at 24 h postoperatively among the four groups.

Discussion
The results of this prospective double-blind randomized study suggest that dexmedetomidine loading dose 0.5 μg/kg and maintenance dose 0.5 μg/kg/h could increase serum potassium concentration slightly during perioperative period.
The incidence of hypokalemia resulted from dexmedetomidine when administered for long-term sedation in the intensive care unit setting was reported to be more than 2% in adult[6] and more than 9% in newborns [7], the mechanism of which is still uncertain. Görges et al. [4] reported that potassium levels were depressed relative to baseline at 15 min and 30 min after intravenous injection of dexmedetomidine as a bolus administered over 60 seconds after general anesthesia induction. However, different from the results of the previous studies, dexmedetomidine did not produce a decreased effect on potassium levels in our study. We found an increase in serum potassium concentration at surgery ending, 5 min after extubation and 24 h postoperatively. when dexmedetomidine loading dose 0.5 μg/kg and maintenance dose 0.5 μg/kg/h was used. The mechanism by which dexmedetomidine causes a mild increase in serum potassium is still unclear.
Catecholamine-induced vasoconstriction results in peripheral tissue ischemia, which causes anaerobic metabolism and lactate production in peripheral tissue [8]. Dexmedetomidine can reduce the sympathetic nervous system activity and is useful in diminishing catecholamine release through its anti-sympatholytic properties [9]. Therefore, dexmedetomidine may alleviate excessive catecholamine-induced lactate overproduction [10,11]. A previous study reported the impact of dexmedetomidine on tourniquet-induced systemic effects in total knee arthroplasty, in which lactic acid concentration was signi cantly lower in the dexmedetomidine group compared with the control group at 10 min after tourniquet release. The authors concluded that the reduced serum lactate level might be related to the reduction of tourniquet-induced pain and hemodynamic uctuations, as well as the regulation of metabolism and coagulation induced from dexmedetomidine. Another study suggested that dexmedetomidine may ameliorate liver damage and decrease lactate level [12]. In the current study, serum lactate level was signi cantly lower in groups D 2 and D 3 at 1 h after surgery beginning compared with the basic value, suggesting that dexmedetomidine loading dose 0.5 μg/kg or 1 μg/kg and maintenance dose 0.5 μg/kg/h might exert certain improvement effects on microcirculatory perfusion.
A transient increase in blood pressure in a short period of time after dexmedetomidine use may be associated with its direct activation on alpha 2 -adrenergic receptors on vascular smooth muscle cells and leading to vasoconstriction, after which blood pressure drops with dexmedetomidine acting on the locus ceruleus, inhibiting the sympathetic nervous system excitation and also increasing the vagal excitability [13,14]. In the current study, it is noteworthy that MAP was signi cantly higher in group D 1 after dexmedetomidine loading dose nish compared to the control group, whereas there were no signi cant changes in MAP in group D 2 at this timepoint, which may be related to the higher basic MAP in the group D 1 than group D 2 . Dexmedetomidine loading dose administration caused further vasoconstriction, resulting in an increase of MAP. MAP and HR in dexmedetomidine groups were signi cantly lower than those in the control group at several time points during operation, at extubation and at 1 hour after into PACU, which might be related to the inhibitory effect of dexmedetomidine on sympathetic nerve.
The antinociceptive effects of dexmedetomidine occurs at dorsal root neuron level, where it blocks the release of substance P in the nociceptive pathway and through action on inhibitory G protein [15,16]. We found that remifentanil dosage in each dexmedetomidine group was signi cantly lower than that in the control group, which might be related to its analgesic effect.
Dexmedetomidine inhibits arginine vasopressin release from the isolated hypothalamic paraventricular nucleus and paraventricular magnocellular neurons and thus producing a diuretic effect [17][18][19]. As reported in the current study, the urine volume in groups D 2 and D 3 was signi cantly higher than that in group C. While there were no signi cant changes in the urine volume in group D 1 , which indicating that lower dose of dexmedetomidine might have no obvious diuretic effect. Though the blood urea nitrogen 24 h postoperatively in group D 1 was signi cantly higher than that before surgery, it was still in normal range, and the renal function in the other groups did not change signi cantly. Previous studies by Jiang et al. [20] have con rmed that dexmedetomidine has a protective effect on the liver. The activity of ALT and AST increased at 24 h postoperatively, which might be related to the decrease of the activity of digestive enzymes in combination with the decrease of nutrient digestibility due to the changes on the intestinal environment. While the ALT and AST levels were all in normal ranges. Additionally, there were no signi cant differences in length of hospital stay, intraperitoneal drainage volume and pulmonary infection incidence 24 h postoperatively among the four groups, which may be related to the modes of operation, the antibiotic treatment and standard procedure of the hospital.
There are still some shortcomings in this study. First, only patients undergoing radical resection of gastrointestinal malignant tumors were included in this study, whether the results are applicable to patients with other types of surgery still needs further study. Second, the three dosages of the dexmedetomidine were selected based on our previous studies reported in the literature. The optimal intra-operative dose of dexmedetomidine remain to be determined, and a further study is needed. Third, perioperative infusion of 0.9% sodium chloride solution may have a certain impact on electrolyte determination. Finally, randomized controlled trials of large samples and multicenter studies are needed to provide reliable evidence for the use of dexmedetomidine.
In summary, dexmedetomidine loading dose 0.5 μg/kg and maintenance dose 0.5 μg/kg/h can elevate serum potassium concentration slightly, reduce lactic acid and remifentanil use signi cantly, relieve perioperative hemodynamic uctuations, and have no adverse effect on postoperative rehabilitation in patients undergoing radical resection of gastrointestinal malignancy. Competing interests: The authors certify that there was no con ict of interest with any nancial organization regarding the material discussed in the manuscript.

Abbreviations
Availability of data and materials: The datasets used and analysed during the current study are available from the corresponding author on reasonable request.
Authors' Contributions: Zhang Zhuan designed the study, drafted and wrote the manuscript. Shunping Tian, Dongsheng Zhang, Ying Wang and Leyang Yu implemented the trial and contributed samples collection. Shinting Yan and Wei Zhou prepared drugs. Ning Li and Yanlong Yu collected the data and did statistical analysis. All authors reviewed the manuscript.    Data are expressed as mean ± SD. P a , P b and P c for comparison between group C and groups D 1 , D 2 and D 3 , respectively. P d for comparison between Groups D 2 and D 3 . Data are expressed as mean ± SD. P a , P b and P c for comparison between group C and groups D 1 , D 2 and D 3 , respectively. P d for comparison between Groups D 2 and D 3 . Data are expressed as mean ± SD. P a , P b and P c for comparison between group C and groups D 1 , D 2 and D 3 , respectively. P d for comparison between Groups D 2 and D 3 . Data are expressed as mean ± SD. ALT: alanine aminotransferase. AST: aspartate aminotransferase. Consort ow diagram for the study.