Effects and mechanisms of glucose-insulin-potassium on post-procedural myocardial injury after percutaneous coronary intervention

Abstract

Objective

To evaluate the effects and mechanisms of glucose-insulin-potassium (GIK) on post-procedural myocardial injury (PMI) after percutaneous coronary intervention (PCI).

Methods

A total of 200 non-diabetic patients with documented coronary heart disease (CHD) were divided into the Group GIK and Group G, with 100 patients in each group. Patients in Group G were given intravenous infusion of glucose solution 2 hours before PCI. As compared, patients in Group GIK were given GIK.

Results

Both post-procedural creatine phosphokinase isoenzyme MB (CK-MB; 62.1 ± 47.8 vs. 48.8 ± 52.6 U/L, P = 0.007) and cTnI (0.68 ± 0.83 vs. 0.19 ± 0.24 ng/mL, P < 0.001) in Group GIK were significantly higher than those in Group G. In Group G, 9.0% and 4.0% of patients had post-procedural increases in CK-MB 1-3 times and > 3 times, which were significantly lower than those in Group GIK (14.0% and 7.0%, respectively; all P values < 0.01); 13.0% and 7.0% of patients had post-procedural increases in cTnI 1-3 times and > 3 times, which were also significantly lower than those in Group GIK (21.0% and 13.0%, respectively; all P < 0.001). Pre-procedural (10.2 ± 4.5 vs. 5.1 ± 6.3, P < 0.001) and post-procedural rapid blood glucose (RBG) levels (8.9 ± 3.9 vs. 5.3 ± 5.6, P < 0.001) in Group G were higher than those in Group GIK. In adjusted logistic models, usage of GIK (compared with glucose solution) remained significantly and independently associated with higher risk of post-procedural increases in both CK-MB and cTnI levels > 3 times. Furthermore, pre-procedural RBG levels < 5.0mmol/L were significantly associated with higher risk of post-procedural increases in both CK-MB and cTnI levels.

Conclusions

In non-diabetic patients with CHD, the administration of GIK may increase the risk of PMI due to hypoglycemia induced by GIK.

1. Introduction

Post-procedural myocardial injury (PMI) is one of the major complications of percutaneous coronary intervention (PCI), defined as creatine phosphokinase isoenzyme MB (CK-MB) and/or cardiac troponin (cTn) elevation above the 99^th centile upper reference limit (URL).^[1-3] Long-term follow-up studies revealed that elevated CK-MB and cTn significantly increased the risk of adverse cardiovascular events in patients underwent PCI.^[1-6] How to reduce PMI is an ongoing topic of concern in the field of cardiovascular research in recent years. In 1962, Sodi-pallares, et al.^[7] first proposed that the solution of glucose-insulin-potassium (GIK), namely the polarized fluid, could be used to treat myocardial ischemia/reperfusion injury (IRI). Opie, et al.^[8-13] further explained the probable mechanisms of its cardioprotection and proposed the concept of metabolic therapy. However, with a large number of basic and clinical studies, the efficacy of GIK is still controversial and the exact mechanisms are unclear. GIK has been reported to cause hypoglycemic events.^[14-29] We found that mild to moderately decreased fasting plasma glucose levels (≤ 5 mmol/L) might be associated with a relative increase in risk of mortality, especially in patients with acute coronary syndrome (ACS).^[30-32] Also, we have reviewed the effects of hypoglycemia on cardiovascular events from the perspective of physiology and pathophysiology.^[33] This study aimed to evaluate the effects and mechanisms of GIK compared with the solution of glucose on PMI. We hypothesized that the administration of GIK might increase the risk of PMI due to hypoglycemia induced by GIK.

2. Methods

2.1. Study design and patient population

The study protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki as reflected in a priori approval by the Ethics Review Boards of each participating institution. Written informed consent was obtained from each patient. A total of 200 patients with documented coronary heart disease (CHD) were selected from September 1, 2018 to December 31, 2019 in three tertiary medical centers. According to the random number table, all patients were divided into the experimental group (Group GIK) and the control group (Group G), with 100 patients in each group. Inclusion criteria were: (1) age at least 18 years; (2) patients with significant coronary artery obstruction in at least one major vessel (stenosis > 50% in left main or > 70% in any other epicardial artery); (3) all patients receiving successful PCI without significant residual stenosis in the target vessel; (4) without history of diabetes and/or hypoglycemic drugs usage; and (5) hemoglobin A1c (HbA1c) < 6.5% and fasting blood glucose < 7.0 mmol/L before PCI. Patients with hemodynamic or cardiac electrical instability, contraindications for PCI, significant comorbidities, or unable to give informed consent were excluded from the study. Note that inclusion in other clinical trials did not preclude enrollment in this study.

Patients in Group G were given intravenous infusion of 500 mL 10% glucose solution 2 h before PCI. As compared, patients in Group GIK were given intravenous infusion of GIK (10% glucose solution 500 mL + 10% potassium chloride 10 mL + insulin 12 U) 2 h before PCI. All procedures were performed by experienced senior physicians who were qualified for PCI. Rapid blood glucose (RBG) levels were measured immediately before and after PCI in all patients. Throughout this article, any reference to plasma myocardial injury biomarkers levels, including CK-MB and cTn, will pertain to that obtained at baseline (after an overnight fast of at least 8 h within 24 h of admission) and 24 h after PCI. During hospitalization, patients were treated with aspirin, clopidogrel, ticagrelor, low molecular weight heparin (LMWH), statins, beta-blockers, angiotensin-converting enzyme inhibitors (ACEIs), angiotensin receptor blockers (ARBs) and others according to practice guidelines. As is their routine, center staff abstracted demographic, clinical, and procedural data.

2.2. Statistical analysis

All case record form data were entered into Epidata 3.1 databases (Epidata Association) by different people. Analyses were conducted with SPSS statistical software, version 22.0 (IBM Inc). Continuous variables will be recorded as mean ± SD. Categorical variables will be recorded as counts. The difference between groups will be analyzed using the student t-test to compare the mean values of biomarker abnormalities. Categorical variables will be compared between groups using the chi-square test. All statistical tests were two-sided and P-values of < 0.05 was considered to be statistically significant.

3. Results

3.1. Baseline demographic and clinical characteristics

Overall, the mean age of the cohort was 60.6 ± 10.8 years old, and 111 (55.5%) patients were males. There were 24 (12.0%), 11 (5.5%), 99 (44.5%) and 37 (18.5%) patients with the history of previous myocardial infarction, stroke, hypertension and hyperlipidemia, respectively. Among all patients, 8.5% were type A lesions, 52.0% type B lesions, and 39.5% type C lesions. As shown in Table 1, there were no significant differences in all baseline demographic and clinical characteristics between groups.

1.

Baseline demographic and clinical characteristics

Characteristics	Group G (n = 100)	Group GIK (n = 100)	P-value
Data are presented as mean ± SD or n (%). ACEIs: angiotensin-converting enzyme inhibitors; ARBs: angiotensin Ⅱ receptor blockers; HDL-C: high-density lipoprotein cholesterol; LDL-C: low-density lipoprotein cholesterol; LMWHs: low molecular weight heparin; LVEF: left ventricular ejection fraction.
Age, yrs	61.0 ± 10.3	60.2 ± 10.2	0.306
Male	55 (55.0%)	56 (56.0%)	0.724
Hypertension	48 (48.0%)	51 (51.0%)	0.117
Hyperlipidemia	19 (19.0%)	18 (18.0%)	0.562
Prior myocardial infarction	11 (11.0%)	13 (13.0%)	0.193
Prior stroke	6 (6.0%)	5 (5.0%)	0.891
HbA1c, %	5.5 ± 1.1	5.6 ± 1.4	0.448
Creatinine, μmol/L	70.8 ± 20.3	72.4 ± 19.5	0.146
LDL-C, mmol/L	2.6 ± 0.8	2.6 ± 0.9	0.375
HDL-C, mmol/L	1.0 ± 0.4	1.0 ± 0.3	0.571
K+, mmol/L	4.3 ± 0.5	4.2 ± 0.6	0.490
LVEF	58.2% ± 11.3%	56.7% ± 10.8%	0.104
Type of lesions in coronary artery disease
Type-A lesion	9 (9.0%)	8 (8.0%)
Type-B lesion	51 (51.0%)	53 (53.0%)	0.479
Type-C lesion	40 (40.0%)	39 (39.0%)
Aspirin	99 (99.0%)	100 (100.0%)	0.863
Clopidogrel	87 (87.0%)	83 (83.0%)	0.592
Ticagrelor	13 (13.0%)	17 (17.0%)	0.086
LMWHs	51 (51.0%)	46 (46.0%)	0.250
Statins	97 (97.0%)	99 (99.0%)	0.926
β-blockers	76 (76.0%)	80 (80.0%)	0.138
ACEIs	49 (49.0%)	46 (46.0%)	0.373
ARBs	30 (30.0%)	28 (28.0%)	0.406

3.2. Comparison of fasting plasma glucose and rapid blood glucose levels between groups

As shown in Table 2, there were no significant differences in baseline fasting plasma glucose (FPG) levels between Group G (5.3 ± 0.6 mmol/L) and Group GIK (5.3 ± 0.5 mmol/L), P = 0.784. However, pre-procedural (10.2 ± 4.5 vs. 5.1 ± 6.3, P < 0.001) and post-procedural RBG levels (8.9 ± 3.9 vs. 5.3 ± 5.6, P < 0.001) in Group G were much higher than those in Group GIK. Compared with Group G, there were 4 (4.0% vs. 0, P < 0.001) and 3 (3.0% vs. 0, P < 0.001) patients in Group GIK with pre- and post-procedural RBG levels < 5.0mmol/L, respectively. Furthermore, in Group GIK the minimum pre- and post-procedural RBG levels were 4.0 and 3.6 mmol/L, respectively.

2.

Comparison of fasting plasma glucose and rapid blood glucose levels between groups

Characteristics	Group G, n = 100	Group GIK, n = 100	P-value
Data are presented as mean ± SD or n (%). FPG: fasting plasma glucose; GIK: glucose-insulin-potassium; RBG: rapid blood glucose.
Baseline FPG, mmol/L	5.3 ± 0.6	5.3 ± 0.5	0.784
Pre-procedural RBG, mmol/L	10.2 ± 4.5	5.1 ± 6.3	< 0.001
Post-procedural RBG, mmol/L	8.9 ± 3.9	5.3 ± 5.6	< 0.001
Pre-procedural RBG < 5.0 mmol/L	0	4 (4.0%)	< 0.001
Post-procedural RBG < 5.0 mmol/L	0	3 (3.0%)	< 0.001

3.3. Comparison of myocardial injury biomarkers levels between groups

As shown in Table 3, there were no significant differences in pre-procedural myocardial injury biomarkers levels between Group G and Group GIK (CK-MB: 18.0 ± 11.2 vs. 17.9 ± 14.5 U/L, P = 0.336; cTnI: 0.01 ± 0.02 vs. 0.02 ± 0.02 ng/mL, P = 0.483). However, both post-procedural CK-MB (62.1 ± 47.8 vs. 48.8 ± 52.6 U/L, P = 0.007) and cTnI levels (0.68 ± 0.83 vs. 0.19 ± 0.24 ng/mL, P < 0.001) in Group GIK were significantly higher than those in Group G.

3.

Comparison of myocardial injury biomarkers levels between groups

Characteristics	Group G (n = 100)	Group GIK (n = 100)	P-value
Data are presented as mean ± SD or n (%). CK-MB: creatine kinase MB; cTnI: cardiac troponin I.
Pre-procedural CK-MB, U/L	18.0 ± 11.2	17.9 ± 14.5	0.336
Pre-procedural cTnI, ng/mL	0.01 ± 0.02	0.02 ± 0.02	0.483
Post-procedural CK-MB, U/L	48.8 ± 52.6	62.1 ± 47.8	0.007
Post-procedural cTnI, ng/mL	0.19 ± 0.24	0.68 ± 0.83	< 0.001
Post-procedural increase in CK-MB 1-3 times, %	9 (9.0%)	13 (13.0%)	0.008
Post-procedural increase in CK-MB > 3 times, %	4 (4.0%)	7 (7.0%)	0.002
Post-procedural increase in cTnI 1-3 times, %	14 (14.0%)	21 (21.0%)	< 0.001
Post-procedural increase in cTnI > 3 times, %	7 (7.0%)	13 (13.0%)	< 0.001

In Group G, 9.0% and 4.0% of patients had post-procedural increases in CK-MB 1 to 3 times and > 3 times above the URL (25U/L), which were significantly lower than those in Group GIK (14.0% and 7.0%, respectively; all P < 0.01); 13.0% and 7.0% of patients had post-procedural increases in cTnI 1 to 3 times and > 3 times above the URL (0.05 ng/mL), which were also significantly lower than those in Group GIK (21.0% and 13.0%, respectively; all P values < 0.001).

3.4. Association between usage of GIK or glucose solution with post-procedural increases in myocardial injury biomarkers levels

To evaluate the association between usage of GIK or glucose solution with post-procedural increases in CK-MB and cTnI levels, we performed logistic regression analysis. After adjustment of baseline demographic and clinical characteristics (including age, gender, previous myocardial infarction, stroke, hypertension, fasting plasma glucose, HbA1c, creatinine, LDL-C, HDL-C, LVEF, type of lesions in coronary artery disease, and pharmacological treatment during hospitalization) in logistic models, usage of GIK (compared with glucose solution) remained significantly and independently associated with higher risk of post-procedural increases in both CK-MB and cTnI levels > 3 times above the URL (OR for post-procedural increases in CK-MB > 3 times: 1.252, 95% CI: 1.097-1.869, P = 0.018; OR for post- procedural increases in cTnI > 3 times: 1.443, 95% CI: 1.260- 2.794, P = 0.014). Furthermore, usage of GIK tended to, but not significantly, be associated with higher risk of post-procedural increases in cTnI levels 1 to 3 times above the URL (OR: 1.175, 95% CI: 0.974-1.901, P = 0.062).

3.5. Association between glucose levels with post-procedural increases in myocardial injury biomarkers levels

Whether in unadjusted or adjusted models, baseline FPG level was not significantly associated with post-procedural increases in myocardial injury biomarkers levels. After adjustment of baseline demographic and clinical characteristics (including age, gender, previous myocardial infarction, stroke, hypertension, fasting plasma glucose, HbA1c, creatinine, LDL-C, HDL-C, LVEF, type of lesions in coronary artery disease, and pharmacological treatment during hospitalization), pre-procedural RBG levels < 5.0 mmol/L were significantly associated with higher risk of post-procedural increases in both CK-MB and cTnI levels above the URL (OR for post-procedural increases in CK-MB 1-3 times: 1.548, 95% CI: 1.040-3.553, P = 0.048; OR for post-procedural increases in cTnI 1-3 times: 1.705, 95% CI: 1.317- 3.042, P = 0.009; OR for post-procedural increases in CK- MB > 3 times: 2.150, 95% CI: 1.539-3.728, P = 0.002; OR for post-procedural increases in cTnI > 3 times: 2.482, 95% CI: 1.881-4.563, P < 0.001).

4. Discussion

PMI, which range from mild to extreme elevation of cardiac biomarkers, can result from the common complications of PCI such as distal embolisation, side-branch occlusion, coronary dissection and disruption of collateral flow, and IRI, etc.^{[1-3, 34-36]} The incidence of PMI varies from 5% to 40%, depending on which one biomarker is detected and the time point of sampling.^{[1-3, 34, 35]} Compared with CK-MB, cTn is the more sensitive biomarkers of PMI. Levels of cTnI and cTnT will reach the peak value at about 24-h after PCI.^[37] Although the incidence of adverse cardiovascular events during hospitalization in patients with CK-MB elevation was the same as that in control group, the cardiovascular mortality was significantly increased in patients with elevated CK-MB level at a mean follow-up of four years.^[38] Fuchs, et al.^[39] found that cTnI level > 0.45 ng/mL after PCI and simultaneously elevated CK-MB and cTnI levels were both independent predictors for adverse cardiovascular events during hospitalization. The PMI-related higher risk of adverse cardiovascular events may be affected by the following mechanisms: (1) decreased left ventricular function; (2) ventricular arrhythmias via a small reentrant circuit in the ventricle as a result of scar formation; and (3) elevation in cardiac biomarkers indicating diffuse coronary atherosclerosis.^{[40, 41]}

In 1962, Sodi-pallares, et al.^[7] first proposed that GIK could be used to treat IRI. Opie, et al.^[8-13] then further explained the mechanisms of its cardioprotective effects: (1) providing more energy substrate, and promoting the uptake and utilization of glucose by cardiomyocytes with insulin assistance, so as to eventually promote the functional recovery of ischemic myocardium; (2) activating cardiomyocyte Na⁺-K⁺-ATPase to promote the uptake of K⁺, thereby stabilizing the polarization state of cell membrane and reducing the occurrence of arrhythmia. However, the efficacy of GIK is still controversial and the exact mechanisms are not clear. In this study, we demonstrate for the first time an independent, highly significant, and positive correlation between the usage of GIK (compared with glucose solution) with higher risk of PMI, which may be due to hypoglycemia induced by GIK.

It has been known that intensive glycemic control may increase the risk of hypoglycemia threefold in patients with diabetes, which have been overlooked or dismissed for a long time.^[42] As stated by American Diabetes Association (ADA), the barrier of hypoglycemia precludes maintenance of euglycemia over a lifetime of diabetes.^[43] Furthermore, non-diabetic individuals can also experience hypoglycemic events due to iatrogenic or non-iatrogenic factors.^[44-47] A large number of trials tend to suggest that hypoglycemia may in fact increase cardiovascular risks and mortality in either diabetic or non-diabetic patients with CHD.^[48-58] Usually, hypoglycemia is defined as blood glucose level below 3.9 mmol/L (70 mg/dL) according to the ADA.^[33] Whereas, blood glucose level below 3.3 mmol/L (60 mg/dL) was also used to define hypoglycemia in some studies.^[33] We reported that mild to moderately decreasing FPG levels (≤ 5 mmol/L) were associated with a relative increase in risk of all-cause mortality in diabetic or non-diabetic patients with ACS.^[30-32] Multiple mechanisms may be involved in the impact of hypoglycemia on cardiovascular prognosis, including but not limited to hemodynamic changes, electrophysiological effects, prothrombotic, proinflammatory and atherogenic effects.^[33]

In the present study, no severe hypoglycemia, which met the general definitions of hypoglycemia as stated above, occurred in either group. However, both pre- and post-procedural RBG levels decreased significantly in patients infused with GIK compared with glucose solution. Furthermore, there were 4.0% and 3.0% patients in Group GIK with pre- and post-procedural RBG levels < 5.0 mmol/L, which were significantly and independently associated with higher risk of post-procedural increases in myocardial injury biomarkers levels. So far, we have proved the original hypothesis that the administration of GIK may increase the risk of PMI due to hypoglycemia induced by GIK. Because of the small size and open-label design of this study, the results may have some bias and other limitations. It is inconclusive and large randomized controlled trials are needed.

In conclusion, in non-diabetic patients with CHD, the administration of GIK may increase the risk of PMI due to hypoglycemia induced by GIK.

Competing interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Acknowledgments

The report was supported by grants from the Beijing Nova Program (No. Z121107002512053), the Beijing Health System High Level Health Technology Talent Cultivation Plan (No. 2013-3-013), Beijing Outstanding Talent Training Program (No. 2014000021223ZK32), the National Natural Science Foundation of China (No. 81100143), the Beijing Municipal Administration of Hospitals Clinical Medicine Development of Special Funding Support (No. ZYLX 201303), and the National Key Clinical Speciality Construction Project.

Authors' contributions

All authors contributed to the implementation of the research, discussing the results and giving comments on the manuscript. H.Y.D and H.P contributed to the analysis of the results and writing of the manuscript. Y.S.W, Z.Y.J and Z.Y.X contributed to the design, Y.S.W and Z.Y.J reviewed/ edited/proved the manuscript.