Objectives: The goal of this study was to compare the effects of 2 different regimens on blood glucose levels of living-donor liver transplant.
Materials and Methods: The study participants were randomly allocated to the dextrose in water plus insulin infusion group (group 1, n = 60) or the dextrose in water infusion group (group 2, n = 60) using a sealed envelope technique. Blood glucose levels were measured 3 times during each phase. When the blood glucose level of a patient exceeded the target level, extra insulin was administered via a different intravenous route. The following patient and procedural characteristics were recorded: age, sex, height, weight, body mass index, end-stage liver disease, Model for End-Stage Liver Disease score, total anesthesia time, total surgical time, and number of patients who received an extra bolus of insulin. The following laboratory data were measured pre- and postoperatively: hemoglobin, hematocrit, platelet count, prothrombin time, international normalized ratio, potassium, creatinine, total bilirubin, and albumin.
Results: No hypoglycemia was noted. The recipients exhibited statistically significant differences in blood glucose levels during the dissection and neohepatic phases. Blood glucose levels at every time point were significantly different compared with the first dissection time point in group 1. Excluding the first and second anhepatic time points, blood glucose levels were significantly different as compared with the first dissection time point in group 2 (P < .05).
Conclusions: We concluded that dextrose with water infusion alone may be more effective and result in safer blood glucose levels as compared with dextrose with water plus insulin infusion for living-donor liver transplant recipients. Exogenous continuous insulin administration may induce hyperglycemic attacks, especially during the neohepatic phase of living-donor liver transplant surgery. Further prospective studies that include homogeneous patient subgroups and diabetic recipients are needed to support the use of dextrose plus water infusion without insulin.
Key words : Dextrose, Insulin, Recipient, Operation
Living-donor liver transplant (LDLT) is an important treatment option for end-stage liver disease (ESLD). With the introduction of this procedure, the number of patients awaiting liver transplant has decreased in recent years.1 Similar to other major and long-lasting surgical operations, patients who undergo LDLT have a significant risk of hyperglycemia. The hyperglycemic effect of these operations may be stress-induced or persistent,2,3 and patients with preoperative diabetes mellitus experience greater difficulties.2-4
The detrimental effects of hyperglycemia have been well demonstrated in medical and surgical intensive care unit patients and during the perioperative period. Hyperglycemia worsens patient prognosis, particularly after cerebrovascular events and cardiovascular surgery.2,5 The detrimental effects of hyperglycemia in liver transplant patients include poor graft survival and increased numbers of infections.6-8 There is a paucity of data regarding blood glucose regulation in liver transplant patients, especially those who undergo LDLT. The vast majority of studies are abstracts of congress and retrospective studies performed in large centers.9,10
The management of hyperglycemia during the perioperative period is always a challenge to anesthesiologists, surgeons, and intensivists. Over the years and throughout the literature, controversy has surrounded the blood glucose target levels and methods for controlling it. The majority of clinicians have accepted the 2001 Leuven Intensive Insulin Therapy Trial as a landmark in this area.11 In this study, Berghe et al demonstrated that intensive insulin therapy (target blood glucose concentration of 80-110 mg/dL) improves the survival of cardiac surgical intensive care unit patients. This work led to many other controlled and uncontrolled studies in this area, with the majority reporting no difference with respect to mortality.2,12,13 Later, it was demonstrated that tight glucose control is associated with substantial dilemmas, such as hypoglycemia. Perioperative clinicians are well aware that hypoglycemia is more detrimental than hyperglycemia, especially in terms of the risk of neurologic sequelae.2,5 A large multicenter trial (Normoglycemia in Intensive Care Evaluation-Survival Using Glucose Algorithm Regulation, NICE-SUGAR) was recently published and demonstrated absolute increases in mortality and hypoglycemic attacks with intensive insulin therapy. The new target blood glucose level has again been changed to “nostalgic” levels (140-180 mg/dL).14
The liver plays a central role in regulating whole-body metabolism. Thus, ESLD leads to major alterations in glucose metabolism. In addition, complex interactions arise during liver transplant surgery with respect to these metabolic processes. Specifically, removal of the old liver and adaptation of the new liver disturbs glucose control during the intraoperative period.4,15
In this study, we aimed to compare the effects of 2 different regimens (5% dextrose in water and 5% dextrose in water plus insulin) on blood glucose levels during the dissection, anhepatic, and neohepatic phases of LDLT.
Materials and Methods
The study was approved by the Institutional Ethics Committee of the Medical Faculty of Inonu University. All of the protocols conformed with the ethical guidelines of the 1975 Helsinki Declaration. Written informed consent was obtained from all subjects. A total of 120 patients (14-64 years old, American Society of Anesthesiologists score II/III) scheduled for LDLT recipient surgery were enrolled and randomly allocated to the 5% dextrose in water (D5W) plus insulin (lispro, Humulin R, 100 IU/mL; Eli Lilly and Company, Indianapolis, IN, USA) infusion group (group 1, n = 60) or the D5W infusion group (group 2, n = 60) using a sealed envelope technique. The infusion for group 1 comprised 500 mL D5W plus 5 IU insulin and was continuously infused at 100 mL/h. The infusion of group 2 comprised plain 500 mL D5W and was continuously infused at 10 mL/h. The patient flow chart is shown in Figure 1. There was no potassium supplementation in either group. Intravenous methylprednisolone was administered intraoperatively at 250 and 100 mg during the dissection and anhepatic phases according to the Inonu University immunosuppressive protocol. All operations for this study were performed between October 2010 and August 2011. We excluded recipients who had previously undergone solid-organ transplant or had preoperative diabetes mellitus or acute liver failure.
All recipients were anesthetized using standard anesthetic techniques and monitored with devices including continuous 5-lead electrocardiography, pulse oximetry, capnography, a radial arterial line for blood pressure monitoring, nasopharyngeal temperature, and bispectral index monitoring (model A-2000TM, Aspect Medical System, Newton, MA, USA). A pulmonary artery catheter was placed when deemed necessary according to preoperative echocardiography pressure values. The internal jugular vein was used for central venous pressure monitoring. Large-bore antecubital vein catheters or central venous 8/8.5-F introducer sheaths were used for rapid fluid or blood and blood product replacement. Anesthesia was maintained using isoflurane or desflurane and remifentanil as an analgesic with cisatracurium as a muscle relaxant. All patients were mechanically ventilated (Dräger, Cato, Medizintechnik GmbH D-23542, Lübeck, Germany) with air/oxygen (50%/50%) to maintain an end-tidal carbon dioxide partial pressure of 30 to 40 mm Hg. Fluids, blood products, and catecholamines were administered at the discretion of the attending anesthesiologist. The same surgical team performed all recipient operations. After the operation, all recipients were transferred to the intensive care unit. The same anesthesiologist collected all of the recipients’ operative data.
The following patient and procedure characteristics were recorded: age (y), sex (male/female), height (m), weight (kg), body mass index (kg/m2), type of ESLD, Model for ESLD score, total anesthesia time (min), total surgical time (min), and number of patients who received an extra bolus of insulin. The following laboratory data were measured pre- and postoperatively: hemoglobin, hematocrit, platelet count, prothrombin time, international normalized ratio, potassium, creatinine, total bilirubin, and albumin.
Three phases of the recipient surgery were analyzed: the dissection phase, which was from the induction of anesthesia until portal vein cross-clamping; the anhepatic phase, from cross-clamping until unclamping of the portal vein; and the neohepatic phase, which was from portal vein unclamping until the end of skin closure. Blood glucose levels were measured and recorded 3 times during each phase with a calibrated blood glucose meter and glucose test strip (Lever Chek, TD-422; Muenster, Germany) using arterial blood samples. The target blood glucose level was 150 mg/dL. When the blood glucose level of a patient exceeded the target level, an extra bolus of insulin was administered by a different intravenous route (1 IU insulin for each 25 mg/dL increase in blood glucose concentration).
Data were presented as mean values ± SD. For continuous variables, the Mann–Whitney U test was used, and the chi-square or Fisher exact test was adopted for categorical variables. For within-group analysis, the Wilcoxon signed-rank test was used. Differences were regarded as statistically significant when P < .05. All calculations were performed using SPSS 16.0 software (Armonk, NY, USA).
The demographic, surgical, and anesthetic characteristics of the recipients are shown in Table 1. The patients in group 2 were younger than those in group 1 (P < .05). Overall, 75% of patients in group 1 and 10% of patients in group 2 received an extra bolus of insulin (P < .05). The other characteristics were similar between the 2 groups.
The intraoperative hematologic data are shown in Table 2. The mean platelet counts in group 2 were higher than those in group 1 at the end of surgery (platelet 2, P < .05).
The intraoperative biochemical data are shown in Table 3. There were no statistically significant differences between the 2 groups.
The intraoperative blood glucose levels are presented in Table 4 and Figure 2. We observed significantly higher blood glucose levels in group 1 compared to group 2 during the anhepatic and neohepatic phases (P < .05).
Blood glucose levels were significantly different (P < .05) from those measured at the dissection 1 time point in group 1. With the exception of the anhepatic 1 and 2 measurements, blood glucose levels were significantly different (P < .05) as compared with the dissection 1 time point in group 2. We performed intraphase (dissection, anhepatic, and neohepatic) analyses of blood glucose levels in both groups. During each phase, we compared the initial level with the other measurements (eg, dissection 1 with dissection 2 and dissection 1 with dissection 3). The statistical analyses revealed that the only difference was between the neohepatic 2 and 3 measurements in group 1 (P < .05).
No patients developed hypoglycemia (blood glucose < 60 mg/dL). The mean blood glucose levels were 116 to 175 mg/dL in group 1 and 113 to 145 mg/dL in group 2.
A total of 120 LDLT recipients were prospectively enrolled in this study. There was no intraoperative mortality. We found that isolated DW5 infusion was superior to DW5 plus insulin infusion in terms of hyperglycemia management.
Intraoperative hyperglycemia management has been studied for other major operations.12,13,16,17 Gandhi et al examined 400 cardiac surgery patients during the intraoperative period.12 They compared intensive insulin therapy with conventional treatment and found that intensive insulin therapy did not reduce perioperative death or morbidity. The mean glucose levels were approximately 111 mg/dL in both groups.
Cammu et al evaluated 20 patients who underwent tumor hepatectomy.17 They used the modified Atlanta protocol, and the lower and upper intraoperative blood glucose concentrations were set at 85 and 110 mg/dL. Blood glucose levels ranged from 65 to 176 mg/dL, indicating that the modified Atlanta protocol is a safe and efficient therapy.
Sato et al examined 52 patients who underwent major liver resection.13 They compared glucose plus insulin therapy with conventional treatment. In both groups, 27% of the patients were diabetic. The majority of glucose measurements from nondiabetic patients were in the range of 63 to 110 mg/dL in both groups. They concluded that glucose plus insulin therapy more effectively maintained normoglycemia than standard therapy.
Maeda et al examined 30 patients who underwent hepatic resection.16 They continuously measured blood glucose levels using an artificial pancreas. Both diabetic and nondiabetic patients were included in the study. They did not use insulin during surgery, and the Pringle maneuver was used for hepatic resection. Glucose levels immediately and profoundly increased after unclamping the hepatoduodenal ligament. The authors hypothesized that glucose variation after unclamping might involve glycogen breakdown within hepatocytes due to hypoxia.
The negative effects of hyperglycemia in liver transplant have been retrospectively evaluated in numerous studies.6-8,18,19 Ammori et al evaluated 184 diabetic and nondiabetic liver transplant recipients.6 They reported that intraoperative hyperglycemia was associated with an increased risk of postoperative infection and mortality. The authors selected a blood glucose level of 150 mg/dL as the threshold between strict and poor glucose control, which was also the target level in our study.
Park et al evaluated 680 diabetic and nondiabetic liver transplants.7 They began insulin infusion if the patient’s blood glucose concentration was > 200 mg/dL. Severe intraoperative hyperglycemia (blood glucose ≥ 200 mg/dL) was found to be an independent risk factor for postoperative surgical site infection.
Xia et al used a divided insulin dose regimen as an effective treatment for hyperkalemic patients during the dissection phase.18 The total patient population was 717, and 50 of the patients received the divided dose regimen. The authors found that this therapy was useful for preventing increased serum potassium levels. In our study, the patient potassium levels were within the normal range and comparable between the 2 groups.
Wallia et al retrospectively studied 113 liver and 31 kidney-liver recipients.8 The cutoff glucose concentration was 200 mg/dL. They reported that immediate postoperative hyperglycemia was associated with an increased risk of liver allograft rejection. Keegan et al examined 158 recipients and found that early postoperative insulin protocol use resulted in a mean glucose concentration of 149 mg/dL,19 which was comparable with our target level.
There is no universal cut-off for glucose deviations in surgical patients. There is some vigilance for diabetic surgical patients, but the preoperative prevalence rate of undiagnosed glycemic dysfunction is unknown.3 We examined nondiabetic recipients, and the main limitation of this study is the absence of a screening test for poor glycemic control, such as measuring hemoglobin A1c or lactate level.
Another limitation of our study is the difference between the mean ages of the groups. The patients in group 2 were younger than those in group 1. Although we used a randomization process for patient selection, the use of larger patient populations should prevent this limitation in future studies. A final limitation is recipient heterogeneity according to ESLD type. Further studies can overcome this problem by including patients with similar types of ESLD (eg, only viral or metabolic).
There is some paucity of clinical data regarding intraoperative blood glucose regulation in liver transplant recipients. Previous studies were retrospective or designed to evaluate the perioperative period. In addition, the authors studied deceased donor recipients. Despite these limitations, our findings should inform future well-designed studies of intraoperative glycemic control during LDLT.
Patients with ESLD exhibit normal hepatic glucose production. However, these patients demonstrate exaggerated hyperglycemic and hyperinsulinemic responses following glucose challenge, which can induce peripheral insulin resistance. Hypoglycemia only occurs in the setting of acute liver failure.4 We excluded patients with acute liver failure from the present study, and there were no recipients with baseline hypoglycemia.
Progressive hyperglycemia occurs during the dissection and anhepatic phases despite concurrent insulin concentration increases.4 In our study, the recipients in group 1 exhibited a similar pattern, but the anhepatic 1 and 2 measurements were similar to the baseline levels in group 2 patients. We believe that this was due to concurrent insulin injection in group 1 patients; it is possible that the subjects developed resistance due to exogenous insulin administration.
Anesthesiologists who work with liver transplants are familiar with the hyperglycemic effects that occur when portal vein unclamping is followed by an immediate stepwise increase in glucose concentration. The effect of the reperfusion (neohepatic) phase on glucose levels has been demonstrated in previous studies. The sudden hyperglycemia is due to glucose release by the graft liver.4 Our recipients exhibited a similar pattern, but group 2 patients showed more acceptable blood glucose levels. A possible beneficial effect of plain DW5 infusion was again observed in group 2 patients.
Transplant recipients are at major risk for hyperglycemia. Additional investigations of hyperglycemia prevention and management are needed.10 We hope that our study will provide some knowledge in this area.
In conclusion, plain D5W infusion (10 mL/h) may effectively promote safer blood glucose levels as compared with D5W plus insulin infusion (100 mL/h) for LDLT recipients. Continuous exogenous insulin administration may induce hyperglycemic attacks, especially during the neohepatic phase of LDLT surgery. Further prospective studies that include homogeneous patient subgroups and diabetic recipients are needed to clarify this issue.
Volume : 13
Issue : 1
Pages : 294 - 300
DOI : 10.6002/ect.mesot2014.P137
From the 1Baskent University Faculty of Medicine, Department of Anesthesiology
and Reanimation, Ankara; the 2Inonu University Faculty of Medicine, Department
of Anesthesiology and Reanimation, Malatya; the 3Malatya State Hospital,
Department of Anesthesiology and Reanimation, Malatya; and the 4Ersin Arslan
State Hospital, Department of Anesthesiology and Reanimation, Gaziantep, Turkey
Acknowledgements: The authors declare that they have no sources of funding for this study, and they have no conflicts of interest to declare.
Corresponding author: Ender Gedik, MD, Baskent University Faculty of Medicine, Department of Anesthesiology and Reanimation, 06490 Ankara, Turkey
Phone: +90 312 212 6868, ext. 4841
Fax: +90 312 223 7333
Figure 1. Patient Flow Chart
Figure 2. Intraoperative Patient Blood Glucose Levels
Table 1. Recipient Characteristics
Table 2. Recipients’ Intraoperative Hematologic Data
Table 3. Recipients’ Intraoperative Biochemical Data
Table 4. Recipients’ Intraoperative Blood Glucose Concentrations (mmol/L)