The effect of insulin on the heart

Abstract

Positive inotropic effects of insulin were described early after the isolation of insulin from the pancreas but data on the effect of insulin on the heart are conflicting. Systemic insulin administration results in a reduction in circulating free fatty acids and an improvement in myocardial glucose uptake, which causes an efficiency improvement in the myocardial cell. There is strong evidence that insulin administration results in functional improvement in dysfunctional myocardium. (Neth Heart J 2010;18:197-201.)

After the first isolation of insulin from the pancreas in the early 1920s, Visscher and Muller were probably the first to attribute a positive inotropic effect to insulin in an isolated heart preparation.¹ However, data on the effects of insulin on cardiac function and metabolism remain conflicting. Several aspects of the effect of insulin on myocardial function and metabolism have been studied. In this first of a two part article dealing with the effect of insulin on the heart, we discuss the research that focused on metabolism, function and coronary blood flow.

Insulin administration in intact organisms

Research focusing on the effect of insulin on the heart can be divided into research with isolated heart preparations and research with intact organisms (animals or patients). In an isolated heart preparation, or even isolated cardiac muscle strips, simple addition of insulin to the perfusate or medium will cause myocardial exposure to insulin, without effects on the glucose content of the perfusate or medium. In vivo, however, systemic administration of insulin causes a drop in glucose level, followed by a secondary response to restore the blood glucose level, with an increase in norepinephrine and glucagon levels. To prevent this drop in blood glucose levels, insulin administration is combined with infusion of glucose, and potassium is usually added to prevent hypokalaemia caused by simultaneous entry into the cell with glucose.

Infusion protocols with glucose, insulin and potassium can be distinguished in two types: first, protocols that use infusions with a fixed composition and pay less attention to the blood glucose level, as long as it does not become too low (glucose-insulin-potassium (GIK) infusions, e.g. the protocol described by Rogers and co-workers²) and second, protocols which have a target of (high-dose) insulin administration with euglycaemic control, effectuated by a fixed insulin dose and variable glucose infusion rates, resembling the euglycaemic hyperinsulinaemic clamp (EHC) procedure described by DeFronzo.³ Confusingly, terminology is frequently mixed up in the literature, so reports on the effect of insulin on the heart have to be judged carefully for the procedure used.

Effects of insulin on metabolism

In vivo animal studies

During GIK administration, the uptake of free fatty acids (FFAs) was reduced,⁴ and glucose uptake was enhanced,^4,5 or at similar levels as in controls.⁶ Compared with controls, lactate production was higher^5,7 or at similar levels,⁶ and one study even observed increased lactate uptake.⁴ According to one study, the adenosine triphosphate (ATP) derived from lactate production was actually used by the myocardium, as ATP content was not different from controls.⁷ Finally, several authors were not able to show a net effect on myocardial oxygen consumption.^5,7,8

Human studies and metabolic effects of GIK

GIK infusion in humans results in a reduction in circulating FFAs,^9-14 often below the myocardial uptake threshold that has been determined to be between 100 and 200 μmol/l.⁹ As a result, myocardial uptake of FFAs minimises9-11,13-15 and the uptake of glucose and lactate increases.9-11,13,14 Myocardial respiratory quotient (RQ) changes towards 1.0 (in fasting state, when FFAs are predominantly used as fuel, myocardial RQ is around 0.7).^11,13 However, similar to studies in animals, most studies that determined myocardial oxygen consumption during insulin infusion in humans did not observe a change.^11,13-15

Only one study related myocardial oxygen consumption to a clinical determinant of oxygen consumption (polyvinyl acetate (PVA) loops) and observed a reduced or ‘normalised’ oxygen consumption, especially in the unloaded heart, when GIK was administered.¹⁵ Thus, a remarkable energetic advantage was achieved when metabolism was changed towards glucose metabolism by GIK, while contractility, myocardial blood flow and the slope of the PVA relationship remained unchanged.

The observed effects on plasma FFA levels probably cause a transition from myocardial FFA metabolism (the ‘preferred substrate’) towards carbohydrate metabolism, and it can be calculated that carbohydrate metabolism is more oxygen efficient than fatty acid metabolism.¹⁶ Also, cytosolic glycolysis is enhanced with increased production of cytosolic ATP, which is then directly available to ion pumps localised in the cell membrane and the endoplasmatic reticulum, as well as to the cross-bridges on the myofilaments. This is thought to be one of the ways that GIK infusions may protect (post)-ischaemic myocardium and improve function.

Positive inotropic action of insulin in normal and abnormal myocardium

Several groups made similar observations of the positive inotropic effect of insulin, such as Visscher and Müller,¹ for example in dogs^17-19 but also in isolated preparations of piglets and kittens,²⁰ in hearts from diabetic lambs,²¹ in hearts from guinea pigs or rats,²² and in sheep with streptozotocin-induced diabetes.²³ However, others have not been able to reproduce these results.^13,24-27 These conflicting results may be due to the already normal function of myocardium, in which it is more difficult to enhance myocardial function than in already dysfunctional myocardium, e.g. dysfunction caused by ischaemia.

Effects of insulin administration in patients with left ventricular dysfunction

Reports studying the effects of insulin administration on myocardial function in patients with left ventricular dysfunction are summarised in table 1.^28-35

Table 1.

Insulin and cardiac function improvement.

Author	Number and type of subjects	Protocol/modality	Insulin equivalent μU/kg/h	Result
Khoury28	N=30 Chronic LV dysfunction due to CAD	Insulin 160 U/l + glucose 100 g/l + KCl 40 mmol/l at 30 ml/h for 4 hours vs. LDD wall motion improvement (max 40 μg/kg/min)Echocardiography	64†	Wall motion improvement in 62/244 dysfunctional segments with dobutamine and in 54/219 dysfunctional segments with GIK79% agreement between wall motion improvement with GIK and with dobutamineSmall but significant decrease in end-systolic volume with GIKNo change in heart rate or blood pressure
Klein29	N=39LV dysfunction due to (non-) STEMI, 5-8 days after MI	Insulin 100 mU/kg/h, glucose 200 g/l + KCL 40 mmol/l at variable rate vs. LDD 10 μg/kg/min and placebo (6 controls receiving NaCl instead of GIK)Echocardiography	100	Wall motion improvement in 88/203 dysfunctional segments with dobutamine and in 98/198 dysfunctional segments with GIK93% agreement between wall motion improvement with GIK and with dobutamineNo change in heart rate or blood pressure
Yetkin30	N=32 Recent anterior wall MI (7±2 days old)	Insulin 100 μU/kg/h, glucose 250 g/l + KCl 80 mmol/l at 1 ml/kg/h‡ vs. LDD 10 μg/kg/minEchocardiography	0.1 (100?)‡	LDD: improvement in wall motion in 57/285 segmentsGIK: improvement in wall motion in 62 of 288 segmentsAgreement 96% on patient levelSimilar improvement in wall-motion indexDuring GIK: increase in blood pressure, heart rate and RPP similar to LDD
Yetkin31	N=21proven CAD and reduced LV function	Insulin 100 μΥ/kg/h, glucose 250 g/l + KCl 80 mmol/l at 1 ml/kg/h‡ vs. LDD 10 μg/kg/minEchocardiography	0.1 (100?) ‡	LDD: improvement in wall-motion in 44/265 dysfunctional segmentsGIK: improvement in wall-motion in 52/265 dysfunctional segmentsAgreement on patient level 95% During GIK: increase in blood pressure, heart rate and RPP similar to LDD
Cottin32	N=12LV dysfunction and stable CAD (MI > 6 months old)	Insulin 300 U/l + glucose 300 g/l + potassium 6 g/l at 1 ml/kg/h for 20 minutesEchocardiography	300	Improvement in wall motion score index at 20 minutes with further improvement at 40 and 60 minutesReduction in end-systolic dimension and improvement in EF and fractional shortening at 40 and 60 minutes (but not at 20 minutes). End-diastolic volume unchanged.
Marano33	N=21 patients with recent MI	Insulin 40 U/l + glucose 100 g/l + KCl 80 mmol/l at 1 ml/kg/h for 24 hours (12 patients) vs. saline 0.9% at 1 ml/kg/h (9 patients)Tetrofosmin SPECT	40	Wall motion improvement in 22/155 segments with GIK, 1/144 with saline
Sasso34	N=4422 diabetics, reduced LV function22 normal subjects	Insulin 100 mU/kg/h + glucose 0.72 ml/kg/h for 2 hours vs. saline 2 ml/kg/h for 2 hoursGated blood pool scans (rest and stress)	100	Saline: no effect on EF at rest or at stressGIK: increase in rest EF in diabetics and increased EF with exercise in diabetics and in healthy volunteers
Alan35	N=30Stable CAD and EF <40%	Insulin 300 U/l + glucose 300 g/l + potassium 6 g/l at 1 ml/kg/h for 24 hoursEchocardiography, SPECT scintigraphy	300	Increase in EF (32±8% to 43±12%, p<0.01)Prolongation in diastolic filling periodDecrease in pulmonary capillary wedge pressure

†Calculated for a subject of 75 kg. ‡ Probably erroneously reported as 100 μU/kg/h instead of 100 mU/kg/h. CAD=coronary artery disease, EF=ejection fraction, GIK=glucose-insulin-potassium, LDD=low-dose dobutamine, LV=left ventricle, MI=myocardial infarction.

The results of these studies strongly suggest that GIK infusion enhances left ventricular function, both in patients with recent myocardial infarction as well as in patients with chronic myocardial infarction. Importantly, the studies reporting improvement in left ventricular or segmental myocardial function used high-dose insulin infusion. Furthermore, insulin administration seems to produce a similar enhancement of segmental function as low-dose dobutamine infusion.

Effects of high-dose insulin administration in patients who underwent cardiac surgery

Hiesmayr and co-workers compared low-dose dobutamine infusion with GIK infusion and observed that GIK administration led to an increase in cardiac index in this patient group; the effect was significant compared with baseline but smaller than the increase obtained by administering low-dose dobutamine.³⁶ Whole-body oxygen consumption was reduced during GIK infusion.³⁶ Studies by others suggested less need for inotropic support and higher cardiac index during GIK infusion and faster weaning from intra-aortic balloon pumping after surgery.^37,38 A reduction in mortality was reached in refractory heart failure patients.³⁹ Also in patients with diabetes mellitus, the need for inotropic support was reduced and higher cardiac index was observed.⁴⁰ No difference in the occurrence and size of myocardial damage resulting from cardiac surgery was observed in two studies and cardiac index was the same as in saline infusion in one.^41,42 Thus, it was suggested that there is a positive (inotropic) effect from high-dose insulin administration in patients who underwent cardiac surgery, but this has to be confirmed in larger, possibly randomised trials.

Coronary blood flow

Data on the effect of insulin administration on coronary blood flow are limited. Increased coronary sinus flow during GIK administration was determined in a number of studies.^11,43,44 The observed increases in blood flow are relatively small (10 to 20%); however, and this small increase was confirmed in a study by Iozzo⁴⁵ in healthy humans using PET determined blood flow. It was also observed that insulin augmented the adenosine-induced blood flow increase by 25% in low-dose and 40% in high-dose insulin administration.⁴⁶ However, Ferrannini and co-workers¹³ did not observe an effect of insulin administration on coronary blood flow, determined with direct measurements in the coronary sinus during euglycaemic hyperinsulaemic clamp.

Summary

Insulin administration reduces free circulating FFAs and enhances myocardial glucose uptake, especially at high doses. Probably, high-dose insulin infusion enhances myocardial function significantly, especially in patients with myocardial dysfunction due to coronary artery disease.