Comparison of glucose control by added liraglutide to only insulin infusion in diabetic patient undergoing cardiac surgery: A preliminary randomized-controlled trial

Background: Liraglutide, glucagon-like peptide-1 (GLP-1) receptor agonist, has been investigated for safety and effectiveness for blood glucose (BG) control in a surgical setting. However, there are only a few studies specific to cardiac surgery patients.
Aims: To primarily compare perioperative 1) BG and 2) glycemic variability (GV) between added liraglutide and only insulin infusion in diabetes mellitus (DM) patients undergoing cardiac surgery.
Setting and Design: A randomized control trial was conducted in DM patients undergoing cardiac surgery with cardiopulmonary bypass (CPB). Inclusion criteria were age 20–80 years and DM Type 2.
Material and Methods: The recruited patients were randomly assigned to Group 1 (added liraglutide with insulin infusion) and Group 2 (insulin infusion). Insulin infusion was based on institutional protocol. Point of care testing (POCT) glucose was used for the adjustment of insulin and BG analysis. Continuous glucose monitor (CGM) was for GV analysis (using Standard deviation: SD).
Statistics: t-test, Chi-square or Fisher-exact test, or Mann–Whitney U test.
Results: Finally, 60 patients were in our study (Group 1 = 32 vs Group 2 = 28). Perioperative mean BG levels of Group 1 were significantly lower than Group 2 with a mean difference of 15.9 mg/dL. Nine patients (18.7% vs 10.7%, P = 0.384) had BG of 60–70 with mean BGs (109.1 vs 147.9, P = 0.001) in the morning. Thirteen patients (9.4% vs 35.7%, P = 0.025) had BG >180 mg/dL at the 1st operative hour. SDs were increasing, but lower SD of Group 1 were observed at the postoperative period. Mean of SDs at postoperative day 2 were 23.65 vs 32.79 mg/dL, P = 0.018.
Conclusions: Liraglutide added with insulin infusion can attenuate perioperative BG and is beneficial in the aspect of lowering GV together with BG at the postoperative period in DM patients. Liraglutide can be applied in cardiac surgery but a rearrangement of time and dosage should be further investigated.

Keywords: Blood glucose control, cardiac surgery, glycemic variability, liraglutide

How to cite this article:
Sindhvananda W, Poopuangpairoj W, Jaiprasat T, Ongcharit P. Comparison of glucose control by added liraglutide to only insulin infusion in diabetic patient undergoing cardiac surgery: A preliminary randomized-controlled trial. Ann Card Anaesth 2023;26:63-71
How to cite this URL:
Sindhvananda W, Poopuangpairoj W, Jaiprasat T, Ongcharit P. Comparison of glucose control by added liraglutide to only insulin infusion in diabetic patient undergoing cardiac surgery: A preliminary randomized-controlled trial. Ann Card Anaesth [serial online] 2023 [cited 2023 Jan 4];26:63-71. Available from: https://www.annals.in/text.asp?2023/26/1/63/367013

Introduction

Perioperative hyperglycemia has been a common problem, as high as 40%–60%, in diabetes mellitus (DM) patients undergoing cardiac surgery.[1],[2] Unlike noncardiac surgery, specific causes of hyperglycemia in cardiac surgery are cardiopulmonary bypass (CPB)-induced glucose metabolism disruption[3] and hypothermia-induced disturbance of hormone secretion[4] and effect on glucagon secretion by both factors.[4] In other words, the mechanism of hyperglycemia in cardiac surgery is more complicated and makes it harder to control.

Control of blood glucose (BG) by insulin infusion has been a standard method.[5],[6] Fluctuation of BG or high glycemic variability (GV) has been proved for causing more endothelial injuries, morbidity, and mortality than a consistent level of hyperglycemia.[7],[8],[9] BG control protocol in perioperative cardiac surgery is efficient for lowering BG, but its effect on the variability of BG has not been yet elucidated.[10]

Liraglutide, glucagon-like peptide-1 (GLP-1) receptor agonist, has been proved its safety and effectiveness in lowering BG and GV in DM outpatient settings for many years.[11],[12] There are a few studies that have demonstrated its effectiveness in perioperative glycemic control.[13],[14],[15] Moreover, there is only one study of Hulst AH, et al.[15] that investigated its effectiveness in cardiac surgery.

Thus, the primary objectives of our study were to compare perioperative 1) BG and 2) GV between added liraglutide and only insulin infusion in DM patients undergoing cardiac surgery. Secondary objectives were to compare adverse effects of combination therapy, the side effect of liraglutide, cardiovascular outcomes, and ICU stay.

Methods

Patient recruitment

The present study was a randomized control trial following CONsolidated Standards of Reporting Trials (CONSORT) guideline. After institutional review board (IRB) approval (November 19, 2019), the trial was conducted in DM patients who were set for cardiac surgery undergoing cardiopulmonary bypass (CPB) at the Cardiac Center, King Chulalongkorn Memorial Hospital. Inclusion criteria were 20–80 years of age, DM Type 2 (T2DM), and scheduling for elective valvular heart surgery (VHS) or coronary artery bypass graft (CABG). Exclusion criteria were 1) DM Type 1, 2) insulin-dependent T2DM, 3) BG <60 or >300 mg/dL from 6 pm of the day before surgery, 4) preoperative administration of insulin, glucose, or dextrose solution, 5) preoperative inotropes/vasopressors infusion or mechanical cardiovascular support devices, 6) history of postoperative nausea or vomiting (PONV), 7) thyroid cancer or endocrine neoplasia syndromes, 8) chronic pancreatitis or previous surgery of pancreas, 9) recent steroid administration, 10) pregnancy, and 11) current treatment with GLP-1 analogs. Written informed consent was obtained from all the enrolled samples.

Sample size and randomization

Based on a previous study,[16] the calculated sample size was 28 in each group (mean BG and standard deviation of placebo = 11.1, 2.8 mmol/L; and with added liraglutide = 9.0, 2.1 mmol/L, Type I error = 0.8, Type II error = 0.2).

Recruited patients were randomly assigned by team researchers into Group 1: added liraglutide to BG control with insulin infusion, and Group 2: BG control with insulin infusion by block randomization with variable random computer-generated blocks of four, six, or eight, allocation ratio of 1:1.

CGM installment and data collection

In the evening (5–6 pm) of the day before surgery, team researchers approached patients for informing and attaching CGM sensor (Enlite, iPro2 system, Medtronic®). The sensor was placed by Enlite serter at the lower abdominal area or other recommended area.[17] The sensor was removed after the 3rd postoperative day. The data were retrieved by docking the sensor with iPro2 digital recorder and connecting to Care link iPro website.

Perioperative blood glucose control

Preoperative glycemic management

Based on institutional protocol, DM patients were prepared at preoperative admission period by 1) endocrinologist consultation if random BG >250 or <70 mg/dL on admission, 2) nothing per oral (NPO) order after midnight for case scheduled at 8.30 am, 3) discontinuation of oral hypoglycemic drugs (OHG) in the morning of the surgery day, and 4) insulin would be given if POCT-glucose >300 mg/dL in the morning of the surgery day, and 5) 5% dextrose solution infusion if surgery delay after 10 am.

Perioperative BG monitor

BG monitoring was 1) Point-of-care testing (POCT-glucose: every 4–6 h at preoperative and postoperative periods and every 30–60 min at intraoperative period) and 2) Continuous glucose monitor (CGM). The BG data from POCT-glucose (Dextrostrix, Accucheck Performa®) were used for BG control.

BG control with insulin infusion (institutional perioperative BG control protocol)

We checked POCT-glucose every 30–60 min intraoperatively. We started rapid insulin (RI) infusion at 2 or 3 Units/h according to the first POCT-glucose (before induction of anesthesia) 180–199 or 200–300 mg/dL, respectively. Then, we adjusted infusion rate (institutional OR sliding scale) as follows: RI infusion rate of 1, 2, 3, 4, and 5 Units/h for the consecutive POCT-glucose of 120–140, >140–180, >180–200, >200–250, and >250 mg/dL. Additionally, we gave RI 5 Units intravenously for once POCT-glucose >250 mg/dL. Aim of BG control was within 140–180 mg/dL.

Liraglutide protocol

We administered Liraglutide (Victoza®) four times as follows: First: Liraglutide 1.2 mg subcutaneously in the evening (about 6–7 pm) of the day before planned surgery, Second: Liraglutide 1.2 mg subcutaneously in the morning (about 7–8 am) of surgery day, Third: Liraglutide 1.2 mg subcutaneously in the morning (about 7–8 am) of the 1st postoperative day (POD1), and Forth: Liraglutide 1.2 mg subcutaneously in the morning (about 7–8 am) of the 2nd postoperative day (POD2).

Liraglutide was ordered in a separate order sheet and given to the ward or ICU nurse. Anesthesiologists who performed anesthesia were blinded to this order sheet.

If patients had severe nausea or vomit after the first dose of liraglutide, they would leave the study and would be monitored by POCT-glucose every 4–6 h for the next 12 h and applied institutional perioperative BG control if abnormal BG had occurred.

Intraoperative fluid management and hypoglycemia protocol

We used 0.9% saline or acetar for standard fluid management in cardiac surgery. Hypoglycemia protocol was as follows: POCT-glucose <70 mg/dL administered 50% glucose 5 mL, and POCT-glucose <60 mg/dL administered 50% glucose 10 mL. None of the other dextrose solutions was allowed except in form of the solvent of cardiovascular drugs.

BG control in ICU

Insulin infusion was continued for 24–36 h of ICU admission. Insulin rates were adjusted based on the institutional ICU sliding scale (RI infusion rate of 0, 1, 2, 3, and 4 Units/h for BG of 120–140, >140–180, >180–200, >200–250, and >250 mg/dL, respectively). OHG was reinstituted routinely at ward when patients were capable to orally intake.

Anesthetic management

The patients received their usual cardiac medications in the early morning on the day of surgery. Upon arriving OR, the patients were premedicated with midazolam 0.02 mg/kg and fentanyl 1 mcg/kg. A five-lead EKG, pulse oximetry, and noninvasive blood pressure monitoring were initiated. Then, we inserted a catheter into the radial artery under local anesthesia for invasive blood pressure monitoring. General anesthesia induction consisted of fentanyl 5–10 mcg/kg, midazolam 0.2–0.4 mg/kg, and pancuronium 0.1–0.15 mg/kg. Additionally, propofol 0.5–1 mg/kg was administered as appropriate. After intubation, we inserted a right internal jugular multilumen central venous catheter. Maintenance of anesthesia was with sevoflurane 1%–2%, adjusted by clinical conditions and pancuronium as needed.

Perioperative variables

We recorded demographic data (age, gender, body weight, height, and underlying diseases), cardiac risk index, preoperative left ventricular ejection fraction (LVEF), basic laboratory data, and cardiac medications and regarding diabetes, we itemized fasting blood sugar (FBS), HbA1c, and OHG. We also collected data during the intraoperative period in terms of duration of operation, CPB time, blood pressure, heart rate, and electrolyte imbalance.

Primary outcome variables

We retrieved BG data from CGM serially every 5 min and placed to group 1) preoperative period (Preoperative night: 6 pm–6 am of the day before surgery), 2) the morning day of surgery (morning of operation day: 6–8 am), 3) intraoperative period before CPB (Pre CPB), 4) CPB period (In CPB), 5) post CPB period (Post CPB: after off-CPB time to end of the operation), 6) postoperative period the day of surgery (POD 0: 6 pm–12 pm day of surgery), 7) postoperative period day 1 (POD1: after midnight of the day of surgery to midnight of another day), 8) postoperative period day 2 (POD2: after midnight of POD1 to midnight of another day). Overall perioperative period (Perioperative period) was counted from 6 pm of the day before surgery to midnight of POD2. All BG values from CGM regard to nine periods were analyzed into a mean with a 95% confidence interval.

GV measure in the present study was standard deviation (SD) calculated by using 5-mi interval of BG levels from CGM. SD was calculated in terms of nine group periods.

Additionally, we retrieved BG data from POCT-glucose to summarize events of hypoglycemia (<70 mg/dL), hyperglycemia (>180 mg/dL), and displayed box plots in a timeline with labeling administrations of liraglutide.

Secondary outcome variables

For insulin consumption, we recorded intraoperative, POD 0, POD 1, and POD 2 periods.

Immediate postoperative outcomes (ICU stay, myocardial injury/infarction, wound infection, sepsis, stroke, and death) were collected. We used criteria of increasing of postoperative troponin T >20% and >10 times of preoperative level as an indication for myocardial injury and infarction,[18] in which blood samplings were drawn at admission and ICU arrival. Also, we monitored side effects, nausea, and vomiting before induction, POD1, and POD2.

Statistical analysis

We used software SPSS for analyzing and extracting the P value for the collected data. The data were expressed as numbers, percentage, mean, and 95% confidence interval. Comparison of mean, median, and numbers was carried out using Student t-test (independence t-test for between groups), Mann–Whitney U test, Chi-square test, or Fisher exact test as appropriate. A probability value less than 0.05 was considered statistically significant.

Results

We enrolled 83 patients into our study and excluded 12 patients by our criteria on the day before planned surgery as shown in [Figure 1]. None of the enrolled patients had BG <60 mg/dL, but two enrolled T2DM patients received insulin treatment based on our preoperative BG control protocol from 6 pm of the day before surgery. On the day of planned surgery, three patients left the study after randomization due to surgery cancellation. One patient was excluded due to POCT-glucose in the morning >300 mg/dL. One patient who received dextrose solution due to surgery delay was excluded. CGM accidentally came off in two patients during sterile surgical drape removal at the end of surgery. At the end of the BG monitor, CGM malfunctioned in four patients. Finally, total analyzed subjects were 60 (Group 1 = 32, Group 2 = 28).

Compared demographic data are shown in [Table 1]. There was no significant difference between two groups except for age (68.1 vs 63.2, P = 0.015). DM control method, type of OHG, and outcome of previous BG control in aspects of FBS and HbA1C were not significantly different between the two groups. Conditions based on cardiac risk index, type and duration of operation, and CPB time were comparable. Previous medications (digoxin, angiotensin-converting enzyme inhibitors (ACEI)/angiotensin-receptor blockers (ARBs), calcium channel blockers, diuretics, and acetylsalicylic acid (ASA)/clopidogrel) were compared and revealed no statistical difference.

We demonstrated BG means of nine periods in [Table 2] and found significantly lower in Group 1 at three periods, morning of the operation day (P = 0.001), pre-CPB (P = 0.01), and perioperative period (P = 0.005). The perioperative mean difference was 15.9 mg/dL. The serial box plots of two groups demonstrated lower BG after liraglutide administrations [Figure 2]. Like the analyzed data from CGM, significant differences were found in the morning of the operation day and pre-CPB. Box plots showed a more narrow range between 25th and 75th percentile values in the liraglutide-added group at every period except for during CPB.

Figure 2: Variation of BG and administrations of liraglutide by timeline. Remark: sign 'I' at timeline = 1.2 mg of liraglutide administration (1st = 6-7 pm of day before planned surgery, 2nd= 7-8 am at surgery day, 3rd =7-8 am of POD1, and 4th = 7-8 am of POD2), *,**= statistically significant differences of median between group, Ο (black or white) = out of 1.5 interquartile range

Click here to view

Regarding GV, data, in general, showed an increase in SD from preoperative to postoperative periods [Table 3]. Lower of SDs of Group 1 were statistically significant at preoperative night (P = 0.046) and POD2 (P = 0.018). On the contrary, the SD of Group 2 during the postoperative period was higher than that of Group 1 but with statistical significance only at POD2. Perioperative means of SD showed high values of both groups without statistical significance.

There was no event of hypoglycemia (BG <60 mg/dL) in this study. BG of 60–70 mg/dL in the preoperative night and the morning of operation day were found in both groups without statistical significance. Hyperglycemia at 1st operative hour was present in 3/32 (9.4%) and 10/28 (35.7%) in Group 1 and 2, respectively (P = 0.025). Insulin consumptions were comparable without statistical significance as depicted in [Table 4].

Table 4: Perioperative hypoglycemia, hyperglycemia and insulin consumption

Click here to view

Postoperative troponin level (850.6 [620.8–1142.2] and 1108.2 [781.2–1494.8] ng/L] and inotropic drugs administration (22/32 and 21/28) were not significantly different (P = 0.236 and 0.592).

Median ICU stays were 41 and 40 h in Group 1 and 2, respectively, without statistical significance (P = 0.905). The number of patients who stayed in ICU over 3 days were not statistically different [Table 5].

Postoperative adverse events were compared and shown in [Table 5]. Reoperation to stop bleeding was performed in 0/32 of Group 1 and 3/28 of Group 2, with no statistical significance (Fisher statistic value = 0.095)

Total eight patients (25%) in Group 1 experienced nausea and two of them vomited and received a single dose of ondansetron. There was no patient who reported nausea or vomiting before the induction of anesthesia.

We followed two patients who were excluded after the first dose of liraglutide. One patient had a surgery delay, and one patient got a cancellation. None of them had POCT-glucose <60 mg/dL for 12 h.

Discussion

Liraglutide is an acylated human glucagon-like peptide-1 (GLP-1) analog. Mechanisms of GLP-1 agonists are mainly stimulating insulin secretion and inhibiting glucagon secretion,[19],[20] and also increasing glucose uptake and glycogen synthesis.[20] Liraglutide is slow absorbed following subcutaneous injection, highly bounds to serum albumin, and has a slow rate of elimination,[20] therefore, it is suitable for once-daily dosing. Liraglutide is approved for glycemic control with a low risk of hypoglycemia as monotherapy and in combination with oral hypoglycemic drugs or insulin in T2DM[14],[15],[21] and T1DM.[22] In addition to glycemic control, the Liraglutide Effect and Action in Diabetes: Evaluation of Cardiovascular Outcome Results (LEADER) trial[23] revealed significant reductions in deaths from cardiovascular causes in T2DM who had received liraglutide for 4 years (n = 4,668) in comparison to placebo (n = 4,672). Thus, myocardium and microvascular protections were proposed to be beneficial effects of liraglutide. A study by Chen et al.[24] demonstrated a significantly lower infarction size in patients with ST-segment elevation who received liraglutide for 3 days after coronary intervention than those who received placebo. Although myocardium protection of liraglutide has not been exactly known, some postulates are 1) its natriuretic and vasodilatory effects,[25] 2) amelioration of inflammation in the process of reperfusion injury,[24] and 3) good control of glycemic variability.[8],[9]

The present study was a preliminary RCT that included only T2DM patients undergoing cardiac surgery. We demonstrated a significant lowering of BG in liraglutide-added than the no-liraglutide group in the morning, pre CPB, and overall perioperative periods (mean difference = 15.9 mg/dL). Also, we revealed a significantly higher number of patients who received only insulin had BG >180 mg/dL in the 1st operative hour. Our results demonstrated the efficacy of liraglutide like the previous study by Hulst AH, et al.[15] using 0.6 mg of liraglutide subcutaneously in the evening of the day before surgery and 1.2 mg after induction of anesthesia showed the effectiveness of lowering BG (mean difference = 0.66 mmol/L). As a result, we could conclude that liraglutide could be used as an adjuvant for lowering BG in DM patients undergoing cardiac surgery.

Hypoglycemia is a major concern in DM treatment, especially when combination therapy is administered. We excluded patients who received insulin treatment any time before surgery and monitored POCT-glucose every 4–6 h after liraglutide administration. None of the patient had BG <60 mg/dL. However, we found a total of nine patients (15%) with BG 60–70 mg/dL in the morning. Based on statistical analysis, we could not demonstrate a significant difference between the liraglutide-added and no-liraglutide groups. This event might simply be explained by an effect of NPO. However, the sample size of 30 per group is not sufficient to discard liraglutide involvement.

We found a significant lower BG of liraglutide-added group in the morning and pre-CPB. We designed the first administration of liraglutide at 6 pm on the day before surgery as liraglutide of 1.2 mg has its peak effect at 8–10 h[22] and based on the study of Hulst AH.[15] Despite severe hypoglycemia did not occur in our study, we should be aware of it due to our limitation of small sample size. Unlike non-DM, most T2DM patients received routine OHG. Therefore, we suggest that the evening dose of liraglutide should be omitted in those who received routine OHG.

Regarding BG level during surgery, we did not demonstrate a significant difference of added liraglutide. Due to uncertain conditions of cardiac surgery such as degree of inflammation, hormonal stimulation, CPB time, and drugs needed,[3],[26] BG control should depend on adjustable drugs like insulin. Our study showed insulin consumption was comparable in both groups. So, it confirmed that liraglutide had no benefit in BG control during cardiac surgery. At POD 0, we noticed lower number of patients who had BG >180 mg/dL in liraglutide group (43.75% vs 67.85%, P = 0.061). Thus, increasing the dose of liraglutide (1.8 mg of liraglutide) in the morning of operation day might improve the significance of BG control during POD 0. However, it is mandatory to study in the future.

By using CGM, we can calculate precise SD which reflects the variation of BG. We demonstrated wider variation (SD >20 mg/dL) at postoperative periods compared with pre and intraoperative periods (SD <20 mg/dL) in both groups. According to a study by Matsumoto S, et al.,[27] liraglutide added with an oral hypoglycemic drug in T2DM had the advantage over insulin in terms of lower SD of BG after 24 h of treatment. We found the same benefit of added liraglutide in postoperative GV control, especially on POD2. [[Figure 2], SD 23.65 vs 32.79 mg/dL, P = 0.018].

The other major side effect of liraglutide is nausea and vomiting which is dose-dependent. It related to liraglutide >2.4 mg once daily and occurred as high as 40%.[28] The present study demonstrated the incidence of 25% in the postoperative period in spite of using dose 1.2 mg of liraglutide. A single dose of ondansetron could resolve symptoms for all our patients. Based on Matsumoto S, et al.,[27] liraglutide of 0.9 mg leaded to good tolerance in long-term therapy. Therefore, we may decrease the dose of liraglutide to 0.9 mg for POD 1 and POD 2 to reduce nausea and vomit but remaining good control of GV. However, further study is mandatory.

Long-term liraglutide has been investigated for good cardiovascular outcomes.[12] Short-term liraglutide has revealed its myocardial protection even in acute myocardial ischemia after cardiac surgery.[24] Hulst AH, et al.[21] demonstrated that left ventricular function assessed by echocardiography was better preserved in the liraglutide group than the placebo group. However, the present study could not demonstrate significant differences in myocardial injury and infarction based on troponin criteria between the two groups. Based on calculating by using our results, a sample size of 70 patients per group would be adequate for investigating these secondary outcomes.

The present study was superior regarding liraglutide in cardiac surgery. The advantages were as follows. Firstly, the present study was the first RCT study that included only noninsulin-dependent DM. Secondly, we recruited much more DM patients than the study of Hulst AH, et al.[15] that recruited 42 DM patients and 219 non-DM patients. Third, we followed BG not only intraoperative but also 48 h of the postoperative period. Forth, our exclusion criteria were designed to avoid confounding, for example, limitations of BG range, and treatment by insulin or glucose. Fifth, the present study was a priority to reveal glycemic variability in cardiac surgery and used CGM data for analysis of GV that brought about the precision of BG variation.[29] Sixth, we demonstrated BG and GV in time periods and using data both from POCT-glucose and CGM. Our results showed data analysis from POCT-glucose was correlated with data analysis from CGM. Lastly, we intended to blind in-charge anesthesiologist as administration of liraglutide were done at ward or ICU. This blinding might lead to less bias of BG control.

Our study limitations were 1) the sample size of 60 was too small for detecting the difference of adverse events, 2) postoperative cardiovascular instability is a confounding of postoperative BG control, which it should be recorded or do subgroup analysis, 3) different types of preadmission OHG taking were incapable to control, and 4) six CGM sensors (10%) that did not record BG and accidentally came out off, also CGM data came from interstitial fluid, might lead to some degree of bias and errors.

Conclusions

Liraglutide added with insulin infusion can attenuate perioperative BG and is beneficial in the aspect of lowering GV together with BG at the postoperative period in DM patients. Liraglutide can be applied in cardiac surgery, but a rearrangement of time and dosage should be further investigated.

Acknowledgements

This paper was supported by Ratchadapisek Sompoch Fund (RA 61/134), Faculty of Medicine, Chulalongkorn University.

Declaration of patient consent

The authors certify that they have obtained all appropriate patient consent forms. In the form, the patients have given their consent for their images and other clinical information to be reported in the journal. The patients understand that their names and initials will not be published and due efforts will be made to conceal their identity, but anonymity cannot be guaranteed.

Trial registration:> TCTR 20190118012.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.

References

1.Lazar HL. Hyperglycemia during cardiac surgery. J Thorac Cardiovasc Surg 2006;131:11-3.

2.Sindhvananda W, Surit M. Variations of perioperative cardiac surgery blood glucose and glycemic variability in diabetes and non-diabetes. J Med Assoc Thai 2020;103:1-8.

3.Najmaii S, Redford D, Larson DF. Hyperglycemia as an effect of cardiopulmonary bypass: Intra-operative glucose management. J Extra Corpor Technol 2006;38:168–73.

4.Lehot JJ, Piriz H, Villard J, Cohen R, Guidollet J. Glucose homeostasis. Comparison between hypothermic and normothermic cardiopulmonary bypass. Chest 1992;102:106–11.

5.Lazar HL, McDonnell M, Chipkin SR, Furnary AP, Engelman RM, Sadhu AR, et al. The Society of Thoracic Surgeons practice guideline series: Blood glucose management during adult cardiac surgery. Ann Thorac Surg 2009;87:663-9.

6.Stamou SC, Nussbaum M, Carew JD, Dunn K, Skipper E, Robicsek F, et al. Hypoglycemia with intensive insulin therapy after cardiac surgery: Predisposing factors and association with mortality. J Thorac Cardivasc Surg 2011;142:166-73.

7.Navaratnarajah M, Rea R, Evans R, Gibson F, Antoniades C, Keiralla A, et al. Effect of glycaemic control on complications following cardiac surgery: Literature review. J Cardiothorac Surg 2018;13:10.

8.Besch G, Pili-Floury S, Morel C, Gilard M, Flicoteaux G, Du Mont LC, et al. Impact of post-procedural glycemic variability on cardiovascular morbidity and mortality after transcatheter aortic valve implantation: A post hoc cohort analysis. Cardiovasc Diabetol 2019;18:27.

9.Oka S, Deyama J, Umetani K, Harama T, Shimizu T, Makino A, et al. Glycemic variability is associated with myocardial damage in nondiabetic patients with ST-elevation myocardial infarction. Cardiovasc Endocrinol Metab 2018;7:47–53.

10.Frontoni S, Di Bartolo P, Avogaro A, Bosi E, Paolisso G, Ceriello A. Glucose variability: An emerging target for the treatment of diabetes mellitus. Diabetes Res Clin Pract 2013;102:86–95.

11.Bajaj HS, Venn K, Ye C, Patrick A, Kalra S, Khandwala H, et al. Lowest glucose variability and hypoglycemia are observed with the combination of a GLP-1 receptor agonist and basal insulin (VARIATION Study). Diabetes Care 2017;40:194–200.

12.Marso SP, Daniels GH, Brown-Frandsen K, Kristensen P, Emann JF, Nauck MA, et al. Liraglutide and cardiovascular outcomes in Type 2 Diabetes. N Engl J Med 2016;375:311–22.

13.Katagiri N, Kigawa I, Yahara Y, Ueda Y, Hamano K, Kaneko S. Liraglutide is a perioperative therapeutic option for patients with Type 2 diabetes that undergo elective surgery. Int J Diabetes Clin Diagn 2016;3:117.

14.Kaneko S, Ueda Y, Tahara Y. GLP1 receptor agonist liraglutide is an effective therapeutic option for perioperative glycemic control in Type 2 Diabetes within Enhanced recovery after surgery (ERAS) protocols. Eur Surg Res 2018;59:349–60.

15.Hulst AH, Visscher MJ, Godfried MB, Thiel B, Gerritse BM, Scohy TV, et al. Liraglutide for perioperative management of hyperglycaemia in cardiac surgery patients: A multicentre randomized superiority trial. Diabetes Obes Metab 2020;22:557-65.

16.Sofizadeh S, Imberg H, Olafsdottir AF, Ekelund M, Dahlqvist S, Hirsch I, et al. Effect of liraglutide on times glycaemic ranges as assessed by CGM for Type 2 diabetes patients treated with multiple daily insulin injections. Diabetes Ther 2019;10:2115-30.

17.Cengiz E, Tamborlane WV. A tale of two compartments: Interstitial versus blood glucose monitoring. Diabetes Technol Ther 2009;11(Suppl 1):S11-6.

18.Thygesen K, Alpert JS, Jaffe AS, Chaitman BR, Bax JJ, Morrow DA, et al. Fourth universal definition of myocardial infarction. J Am Coll Cardiol 2018;72:2231–64.

19.Vanderheiden A, Harrison LB, Warshauer JT, Adam-Huet L, Li X, Yuan Q, et al. Mechanisms of action of liraglutide in patients with Type 2 Diabetes treated with high-dose insulin. J Clin Endocrinol Metab 2016;101:1798–806.

20.Drucker DJ, Nauck MA. The incretin system: Glucagon-like peptide-1 receptor agonists and dipeptidyl peptidase-4 inhibitors in type 2 diabetes. Lancet 2006;368:1696-705.

21.Hulst AH, Visscher MJ, Cherpanath TGV, van de Wouw L, Godfied MB, Thiel B, et al. Effects of liraglutide on myocardial function after cardiac surgery: A secondary analysis of the randomised controlled GLOBE Trial. J Clin Med 2020;9:673.

22.Mader JK, Jensen L, Ingwersen SH, Christiansen E, Heller S, Pieber TR. Pharmacokinetic properties of liraglutide as adjunct to insulin in subjects with Type 1 diabetes mellitus. Clin Pharmacokinet 2016;55:1457-63.

23.Poulter NR, Ravn LS, Steinberg WM, Stockner M, Zinman B, Bergenstal RM, et al. Liraglutide and cardiovascular outcomes in type 2 diabetes. N Engl J Med 2016;375:311–22.

24.Chen WR, Chen YD, Tian F, Yang N, Cheng LQ, Hu SY, et al. Effects of liraglutide on reperfusion injury in patients with ST-segment-elevation myocardial infarction. Circ Cardiovasc Imaging 2016;9:e005146. doi: 10.1161/CIRCIMAGING.116.005146.

25.Bizino MB, Jazet IM, Westenberg JJM, van Eyk HJ, Paiman EHM, Smit JWA, et al. Effect of liraglutide on cardiac function in patients with type 2 diabetes mellitus: Randomized placebo-controlled trial. Cardiovasc Diabetol 2019;18:55.

26.Kuntschen FR, Galletti PM, Hahn C, Arnulf JJ, Isetta C, Dor V. Alterations of insulin and glucose metabolism during cardiopulmonary bypass under normothermia. J Thorac Cardiovasc Surg 1985;89:97-106.

27.Matsumoto S, Yanazaki M, Kadono M, Iwase H, Kobayashi K, Okada H, et al. Effects of liraglutide on postprandial insulin and glucagon responses in Japanese patients with type 2 diabetes. J Clin Biochem Nutr 2013;53:68–72.

28.Lean ME, Carraro R, Finer N, Hartvig H, Lindegaard ML, Rossner S, et al. Tolerability of nausea and vomiting and associations with weight loss in a randomized trial of liraglutide in obese, non-diabetic adults. Int J Obes 2014;38:689-97.