Effects of intermittent and long-term glucose-insulin-potassium infusion in patients with systolic heart failure

Abstract

BACKGROUND

Although single dose and short-term glucose-insulin-potassium (GIK) infusions are known to have positive cardiac effects, the effects of repeated and long-term GIK infusion on left ventricular (LV) systolic function and brain natriuretic peptide (BNP) levels are unknown.

OBJECTIVE

To investigate the effects of repeated and long-term GIK infusion on LV systolic function and BNP levels.

METHODS

Thirty-three patients diagnosed with ischemic cardiomyopathy were included in the study. Patients were divided into two groups: the GIK group (n=19) and the control group (n=14). GIK solutions (1000 mL 20% dextrose, 60 U insulin and 50 mmol/L KCl) were administered at 1 mL/kg/h for 24 h on the first, third and fifth days. The patients were examined by echocardiography at 24 h, one week and one month after the start of treatment. BNP levels were measured before and after GIK infusion.

RESULTS

In the GIK group, baseline ejection fraction (EF) was 29.2±10.3%. After one week, EF elevated to 40.8±10.8% (P=0.001). The EF after one month (37.1±10.9%) was less than the EF in the first week, but it was significantly higher than baseline in the GIK group (P=0.01). However, no significant changes in EF were observed after one week and one month in the control group (P=0.1 and P=0.2, respectively). BNP levels after GIK infusion was significantly lower than baseline level in the GIK group (P=0.01).

CONCLUSION

Intermittent and long-term GIK infusion has beneficial effects on LV systolic function in a short and intermediate amount of time. Decrease in BNP levels may indicate effective GIK treatment. Intermittent and long-term GIK infusion could be a promising treatment option in patients with systolic heart failure.

Glucose-insulin-potassium (GIK) infusion has been shown to have favourable cardiac effects by different mechanisms (1–4). Some of these benefits include a reduced infarct size, an enhanced functional recovery and a 28% decreased rate of mortality in patients with acute myocardial infarction (5). In addition, a cardioprotective effect has been observed in patients who have undergone cardiac surgery with GIK infusion.

Beneficial effects of GIK infusion have also been observed in left ventricular (LV) systolic and diastolic dysfunction (6). Previous heart failure studies show improvements in LV contractile function when doses of GIK infusion were varied. However, in most of these studies, GIK infusion was performed only once and in a short period. Only in one study by Alan et al (7) had long-term GIK infusion improved systolic function. Although beneficial effects were demonstrated in these studies, the duration of this effect and how systolic function changes with intermittent and long-term GIK infusion are still unknown.

Brain natriuretic peptide (BNP) is secreted primarily from the ventricles in response to stretch and increased ventricular wall tension. BNP levels correlate with LV end-diastolic pressure and volume, New York Heart Association functional classification, and extent of reversible myocardial ischemia. BNP level has been accepted as a strong predictor of cardiovascular mortality and morbidity (8–10). Previous data suggest that BNP may guide initiation and titration of heart failure therapy (11,12), but there are no data regarding how BNP levels change with GIK infusion.

In the present study, we investigated the effects of intermittent doses and long-term infusions of GIK on LV systolic function and BNP levels. We also examined the effect of GIK infusion on systolic function in early and intermediate periods.

METHODS

Patient selection

Patients who had an ejection fraction (EF) below 45% and who were diagnosed with ischemic cardiomyopathy (CMP) were included in the study. Patients were divided into two groups: the GIK group and the control group. The GIK group consisted of 19 patients and the control group consisted of 14 patients. A diagnosis of ischemic CMP was made if the patient had LV dysfunction, a history of myocardial infarction or coronary revascularization, or epicardial coronary vessel stenosis greater than 70% in coronary angiography. Patients were excluded if they had acute coronary syndrome in the past six months, diabetes mellitus, atrial fibrillation, renal insufficiency (creatinine concentration greater than 1.5 mg/dL), severe heart failure (Killip class III or IV) and New York Heart Association class IV. Standard medication for heart failure was given to all patients in hospital. Enalapril (5 mg to 20 mg, two tablets daily), an angiotensin-converting enzyme inhibitor, and carvedilol (6.25 mg to 12.5 mg, two tablets daily), a beta-blocker, were given to patients. None of the patients had taken other inotropic agents such as levosimendan, dopamine and dobutamine. The study was approved by a local ethics committee.

GIK protocol

When patients became hemodynamically stable after receiving standard medical therapy, GIK infusion was started in the GIK group. The GIK solution consisted of 60 U soluble human insulin and 50 mmol/L KCl in 1000 mL of 20% dextrose. On the first, third and fifth days, the GIK solutions were administered to patients at 1 mL/kg/h for 24 h. GIK infusion was stopped in patients with worsening clinical or hemodynamic stability, hyperkalemia (greater than 5.5 mmol/L), hypo- or hyperglycemia, and increase in any cardiac marker. In the control group, optimal conventional heart failure therapy was given to all patients (Table 1).

TABLE 1.

Baseline characteristics of study groups

	GIK group (n=19)	Control group (n=14)	P
Age, years	69.9±13.9	67.5±12.1	NS
Sex
Male:Female	11:8	9:5	NS
Male, n (%)	11 (58)	6 (43)	NS
Female, n (%)	8 (42)	8 (57)	NS
BUN, mg/dL	55±17	52±15	NS
Creatinine, mg/dL	1.1±0.5	1.0±0.4	NS
Serum Na+, mmol/L	136±9	139±11	NS
Serum K+, mmol/L	5.1±1.1	4.9±0.9	NS
Heart rate, beats/min	84±17	82±17	NS
Blood pressure, mmHg
Systolic	121±19	118±16	NS
Diastolic	71±12	70±10	NS
Before GIK infusion
Left ventricular EF (%)	29.2±10.3	30.1±7.3	NS
NYHA class	2.3	2.4	NS
Medication used in hospital, %
ACE inhibitor or ARB	100	100	NS
Diuretics	100	100	NS
Nitroglycerine	68	57	NS
Acetylsalicylic acid	78	65	NS
Digoxin	58	65	NS
Beta-blockers	47	50	NS

Data are expressed as mean ± SD unless otherwise indicated. ACE Angiotensin-converting enzyme; ARB Angiotensin receptor blocker; BUN Blood urea nitrogen; CMP Cardiomyopathy; EF Ejection fraction; GIK Glucose-insulin-potassium; NYHA New York Heart Association

Biochemical analyses

Creatine kinase and its MB isoform were measured twice daily with using the Konelab 60i analyzer (Thermo Labsystems, Finland). Troponin I was measured with ELFA, a one-step immunoassay sandwich method, using monoclonal antibodies (bioMérieux Vidas, USA). Any significant increases in these markers were accepted as indicators of myocardial damage. BNP levels (ng/L) were measured before and after GIK infusion (after the first week) with the Triage B-type Natriuretic Peptide test (Biosite Diagnostics, USA).

Echocardiography

The patients were evaluated by echocardiography at baseline, after the first week and after one month. Echocardiographic images were obtained using a Wingmed System V echocardiography device (General Electric, Norway). LV end-diastolic and end-systolic volumes, as well as EF, were calculated using the modified Simpson’s rule in apical views (13). On the parasternal long axis, LV end-diastolic and end-systolic diameters were measured by M mode. All echocardiographic examinations were performed by a cardiologist who was blinded to the subjects’ data.

Data were expressed as mean ± SD or as proportions. Paired t test was used to investigate the time dependent variables in the two groups. P<0.05 was accepted as statistically significant. The SPSS version 11.5 software package (SPSS Inc, USA) was used for statistical analysis.

RESULTS

Baseline characteristics of the GIK and control groups are presented in Table 1. At baseline, the mean (± SD) LV EF was 29.2±10.3% in the GIK group and 30.1±7.3% in the control group. During GIK infusion, an increase in cardiac markers was observed in two patients. Hyperkalemia was found in one patient. These three patients were excluded from the study.

In the GIK group, the mean EF was 40.8±10.8% after the first week (P=0.001). The mean LV EF after one month (37.1±10.9%) was less than the LV EF after the first week, but it was significantly higher than baseline (P=0.01). In the control group, there was no significant change in LV EF from baseline, after the first week and after one month (30.1±7.3%, 33.2±6.0% and 31.1±7.3%, respectively) (Figure 1). While LV end-diastolic diameters one week after GIK infusion were similar to baseline diameters in the GIK group (61.2±3.2 mm versus 61.7±3.3 mm, respectively; P=0.2), LV end-systolic diameters were significantly reduced after GIK infusion (systolic diameter 54.9±6.4 mm versus 51.8±6.8 mm, respectively; P=0.01).

Figure 1.

Measurements of left ventricular (LV) ejection fraction at baseline, after the first week and after one month. Data are expressed as means. P<0.05 was statistically significant. GIK Glucose-insulin-potassium; NS Not significant

In the GIK group, BNP levels decreased in 13 patients and increased in three patients. Mean BNP level after GIK infusion was significantly lower than baseline BNP level (335±101 ng/L versus 516±155 ng/L, respectively; P=0.01). In the control group, mean BNP level was lower than baseline BNP level. However, this difference was not statistically significant (463±85 ng/L versus 493±106 ng/L, respectively; P=0.39).

DISCUSSION

Treatment of heart failure is an important clinical issue. Despite current treatments, there is a great need for new treatment options in patients with congestive heart failure. The present study is the first clinical preliminary study concerning the use of intermittent and long-term GIK infusion in patients with systolic heart failure. We demonstrated that intermittent and long-term GIK infusion has favourable effects in patients with systolic heart failure.

Previous studies (6,7,14,15) have demonstrated that GIK infusion improves systolic and diastolic function, as well as segmental or global wall motion. Cottin et al (14) used single-dose and short-term GIK infusion in patients with chronic ischemic CMP. They found that EF increased after GIK infusion. In the present study, maximal effects appeared 40 min after the complete GIK infusion but any information about the duration of these effects was not apparent. In another study, Alan et al (7) demonstrated that both systolic and diastolic functions, and hemodynamic parameters such as pulmonary artery pressure and pulmonary capillary wedge pressure, were improved with GIK infusion. Although long-term GIK infusion was used in that study, infusion was given only once and the study did not state whether these effects were permanent or temporary. We used long-term and repeated dose GIK infusion, which is different from the previous studies. We suspect that the cardiac effects may be more prominent, and that these benefits continue after long-term and repeated dose GIK infusion. Our results have confirmed these suspicions. In our patients, improvements of systolic functions were more marked than previous GIK studies. Moreover, this favourable effect had continued into the first month.

LV dysfunction in patients with ischemic CMP is often caused by myocardial stunning, hibernation and scar tissue (16). Myocardial stunning is characterized by reversible impaired postischemic systolic and diastolic dysfunction (17). We suspect that stunned and hibernating myocardium may have important roles in the pathophysiology of GIK infusion. The mechanisms responsible for myocardial stunning are not yet fully understood. However, possible mechanisms for myocardial stunning involve the generation of oxygen free radicals and the alteration in contractile protein structure (18,19). Calcium (Ca²⁺) overload, Ca²⁺-ATPase dysfunction and decreased sensitivity to Ca²⁺ may contribute to the development of stunning (20). Although the effects of GIK infusion on myocytes may occur through various mechanisms, such as metabolic effects, direct hemodynamic effects, coronary flow improvements and catecholamine-mediated effects (21–24), GIK infusion on Ca²⁺ homeostasis may have greater benefit in improving systolic function. Consequently, Ca²⁺ is essential for cardiac contraction and Ca²⁺ sensitizing agents are used in promising approaches to treat patients with heart failure. We believe that GIK infusion may improve Ca²⁺ homeostasis through one of its different effects. By effecting Ca²⁺ homeostasis in patients with ischemic and nonischemic CMP, GIK infusion may increase myocardial contractility.

We demonstrated that the levels of BNP diminished after GIK infusion compared with the level before GIK infusion. This decrease in BNP levels may have clinical importance. BNP is a sensitive marker of cardiac dysfunction and is a strong predictor of long-term mortality (10). A previous study (25) suggests that patients with ‘low’ BNP levels (200 ng/L or less) on stabilization may be considered at very low risk of events. Conversely, BNP levels (500 ng/L or greater) on discharge may be considered high risk (25). In another similar study (26), while the rate of cardiac events was 23.5% at one month and 79.4% at six months in patients with predischarge BNP levels of greater than 350 ng/L, this rate was 0% and 12.7%, respectively, for patients with predischarge BNP levels of less than 350 ng/L. These results emphasize our study results concerning BNP levels. We believe that the decreased BNP levels with GIK infusion in our study might be a marker of fewer cardiac events after discharge. In addition, we also believe that the decrease in BNP levels may be associated with the efficiency of GIK therapy and therefore it may be used to evaluate the efficacy of GIK therapy.

The most important side effect of GIK infusion in previous GIK trials was hyperkalemia, which occurred in approximately 10% of patients (27). Similarly, hyperkalemia was observed in only one patient (6%) in our study. However, to our surprise, cardiac markers, which indicate myocardial damage, had increased in two patients (12%) during GIK infusion. There are no data concerning myocardial damage in previous GIK trials. This side effect should be considered in patients receiving GIK infusion. An important limitation in the present study is the limited number of patients. However, our study is only a preliminary study. Further studies are needed to clarify the effects of intermittent and long-term GIK infusion on patients with systolic heart failure.

CONCLUSIONS

The intermittent and long-term GIK infusion has beneficial effects on LV systolic functions in the short and intermittent period. This benefit was prominently seen in the first week and was maintained in the first month with slight decrease. This intermittent and long-term GIK infusion regimen could be a new treatment option in patients with CMP.