At least two-thirds of patients with type 2 diabetes will die from cardiovascular disease, and of these two-thirds will die from the manifestations of ischemic heart disease. The National Cholesterol Education Program (NCEP) recognizes diabetes as a cardiac equivalent, which means that in the next 10 years there is at least a 20% chance (the actual figure is 36%) that a patient with diabetes will suffer a cardiovascular event. This places patients with diabetes in the same category as patients with abdominal aortic aneurysms, symptomatic peripheral vascular disease, or a history of stroke.
At least two-thirds of patients with type 2 diabetes will die from cardiovascular disease, and of these two-thirds will die from the manifestations of ischemic heart disease. The National Cholesterol Education Program (NCEP) recognizes diabetes as a cardiac equivalent, which means that in the next 10 years there is at least a 20% chance (the actual figure is 36%) that a patient with diabetes will suffer a cardiovascular event. This places patients with diabetes in the same category as patients with abdominal aortic aneurysms, symptomatic peripheral vascular disease, or a history of stroke. In fact, a patient with diabetes and no known ischemic heart disease has as much chance of suffering a myocardial infarction (MI) as a non-diabetic patient who has already had an MI.1
When an MI occurs in a patient with diabetes, in-hospital, pre-hospital, and one-year mortality is approximately double that of a non-diabetic MI patient. Included in this poor prognosis is an increased prevalence of ventricular fibrillation, atrial fibrillation, and heart failure (HF).2,3 Indeed, even with a less than ‘full-thickness’ infarct of the ventricular wall (a non-ST segment elevation MI [STEMI]), the prognosis for the patient with diabetes approaches that of the non-diabetic with a full-thickness infarct (STEMI). When a cardiac arrest occurs in a patient with diabetes, the chances of being discharged from hospital are greatly decreased.
Interventions in patients with diabetes are also less successful. Angioplasty with or without stent placement, fibrinolysis, and coronary artery bypass grafting are less efficacious. The re-stenosis rate following stent placement in a patient with diabetes is increased, and while drug-eluting stents (DES) reduce stenosis by two-thirds, re-intervention is still twice as frequent in those with diabetes.4 This is particularly true in those who require insulin and/or have multiple vessel disease. The more deadly complication of stent thrombosis is increased by 80% in those with diabetes, and again this is more common in those who require insulin and/or have multiple vessel disease.5 However, in patients with diabetes, stent placement following an MI results in a decrease in recurrent MI, stroke, and death compared with fibrinolysis alone.
In those with type 1 diabetes below 35 years of age, there is no increase in cardiac events. However, if proteinuria develops at any age, the incidence of cardiac events is increased by as much as 140%. The presence of proteinuria and/or renal decompensation results in increased platelet aggregation, decreased fibrinolysis, oxidative stress, endothelial dysfunction, insulin resistance, lower high-density lipoprotein (HDL) cholesterol, and higher triglycerides and low-density lipoprotein (LDL) cholesterol. The four-fold increase in cardiac events that occurs above 35 years of age in a patient with type 1 diabetes could be explained by the development of microalbuminuria, which is also associated with the same increased cardiac risk factors as proteinuria.
Risk Factors for Cardiac Events in Patients with Type 2 Diabetes
The UK Prospective Diabetes Study (UKPDS) of the newly diagnosed type 2 diabetes patient showed that there were only five significant risk factors for an MI. In order of significance these were elevated LDL cholesterol, decreased HDL cholesterol, increased glycated hemoglobin (HbA1c) and systolic hypertension, and smoking.6 Therefore, at any office visit, all of these risk factors should be prioritized, assessed, and treated.
Lowering Low-density Lipoprotein and Statins
As diabetes is a cardiac equivalent, the goal for calculated LDL cholesterol in a patient with type 2 diabetes should be equal to that of a patient with established coronary artery disease, i.e. at least ≤100mg/dl and preferably ≤70mg/dl. Since LDL particle size is reduced with insulin resistance, the calculated LDL level in a diabetic patient can be misleading. In addition, these smaller LDL particles, especially if glycated, are more atherogenic because they are more easily oxidized and thus bind more easily to the scavenger receptor on the macrophage to participate in formation of the atheromatous plaque. A better measure of cardiovascular risk in the insulin-resistant or type 2 diabetes patient is non-HDL cholesterol (total cholesterol minus HDL cholesterol), which measures all of the atherogenic particles (LDL cholesterol, remnant intermediate-density lipoproteins, and small dense very LDL [VLDL] cholesterol particles). Since statins also have pleotrophic effects (decreased inflammation, improved endothelial function, lowered insulin resistance, and decreased oxidation of LDL cholesterol particles), many physicians and groups advocate statin therapy in all type 2 diabetes patients, irrespective of their calculated LDL.
Raising High-density Lipoprotein
A major goal for the patient with type 2 diabetes and dyslipidemia is to elevate their HDL cholesterol level.7 While this is more rewarding than lowering the LDL cholesterol level, it is also more difficult. For every 1% that the LDL is lowered there is a 1% decrease in cardiac events, whereas for every 1% the HDL cholesterol level is elevated there is a 3% reduction. HDL cholesterol levels are low in the insulin-resistant patient with diabetes due to decreased hepatic HDL cholesterol production, especially in the presence of hepatic steatosis (which is associated with the insulin resistance syndrome) and decreased HDL particle half-life. The decreased longevity of the HDL cholesterol particle is due to the particle being small and more easily metabolized by the liver. Furthermore, these small HDL cholesterol particles are less effective than larger HDL cholesterol particles in improving reverse cholesterol transportation, inflammation, and endothelial function and decreasing LDL cholesterol oxidation. Therefore, lowering insulin resistance with exercise, weight loss, alcohol, and insulin sensitizers is recommended to increase HDL cholesterol levels in those with diabetes. However, the use of nicotinic acid and, to a lesser degree, fibrates is more effective in raising HDL cholesterol levels. In addition, utilizing a statin such as rosouvastatin—which will increase HDL cholesterol levels compared with atorvastatin, which at high doses will decrease HDL cholesterol levels—with atorvastatin is prudent. However, there is an urgent need for the development of drugs that not only increase HDL cholesterol levels but also improve the function of the HDL cholesterol particle.
The third most important risk factor for an MI in those with type 2 diabetes is glycemic control. In the European Prospective Investigation into Cancer in Norfolk (the EPIC-Norfolk) study, the lower the HbA1c level (down to a minimum of 5%), the lower the risk for an event related to coronary artery disease, cardiovascular events, and death. In addition, for every 1% that the HbA1c level rose above 5%, there was a 24 and 28% increase in mortality for men and women, respectively.8 These epidemiological data are similar to the UKPDS, where the closer the HbA1c level came to 5.5%, the lower the risk for an MI.9 Furthermore, in a Danish study of patients with diabetes who had developed HF following a myocardial infarction, for each 1% that the HbA1c level rose above 5%, mortality increased by 24%. Thus, glycemic control as judged by the HbA1c is an important risk factor for cardiac events. However, at levels below 7.3% the majority of the HbA1c level is due to post-prandial rather than fasting or pre-prandial glucose levels.
As shown in the Diabetes Epidemiology: Collaborative Analysis Of Diagnostic Criteria in Europe (DECODE) study, cardiac risk could be predicted from the post-prandial (two hours after eating) glucose on a glucose tolerance test but not from the fasting glucose. Post-prandial hyperglycemia, irrespective of the level of fasting and pre-prandial glycemia, induces oxidative stress to levels that are not obtained with chronic sustained hyperglycemia.10 Post-prandial hyperglycemia is accompanied by post-prandial hyperlipidemia (elevated free fatty acids and triglycerides), and this ‘deadly duo’ of post-prandial hyperlipidemia and hyperglycemia (post-prandial dysmetabolism) through increased cytokine production induces insulin resistance, inflammation, endothelial dysfunction, platelet aggregation, and decreased fibrinolysis due to increased PAI1 levels.11 Post-prandial dysmetabolism also leads to increased inflammation within the atheromatous plaque, which in turn increases the risk for plaque rupture and a cardiovascular event. Post-prandial dysmetabolism leads to increased atheroma formation, and this can be seen in studies where the increase in carotid intima-medial thickening was proportional to post-prandial glucose and lipid levels. In addition, a prospective angiographic study of post-menopausal euglycemic women with coronary artery disease has shown that over a three-year period the lower the post-prandial glucose level (to a level of 87mg/dl), the less the deterioration in the diameter of the coronary artery lumen. Indeed, if the post-prandial glucose was less than 86mg/dl, the coronary artery diameter increased.12
That the treatment of post-prandial dysmetabolism lowers the risk for cardiac events was seen in trials of the α-glucosidase inhibitor acarbose, a drug whose action is confined to lowering post-prandial and not pre-prandial or fasting glucose levels. Acarbose administered to subjects with glucose intolerance reduced progression to diabetes by 25%, but decreased cardiac events by 49%.13 A re-analysis of the phase III placebo-controlled studies of acarbose showed a 64% decrease in MI and a 35% decrease in cardiac events.
Acarbose was also shown to reversibly reduce carotid intimal-medial thickening by 50%.14 In addition to the α-glucosidase inhibitors, drugs that have the ability to control post-prandial glucose include rapid-acting insulins, meglitanides, some sulfonylureas, thiazolidinediones, pramlatide, incretin mimetics, and dipeptidyl peptidase-4 (DPP-4)-inhibitors. In addition to their metabolic effects, these drugs may also play a role in decreasing cardiovascular events.
In patients with or without diabetes, if hyperglycemia develops following an MI, the prognosis is worse. This is because the presence of hyperglycemia closes the myocardial adenosine triphosphate (ATP)-sensitive potassium channels, and this closure results in decreased myocardial ATP levels, increased myocardial ischemia, and an increased risk for arrhythmias, HF, and death. Controlling the stress hyperglycemia with intravenous insulin is a powerful tool in improving the prognosis of the hyperglycemic post-MI patient. Fluctuations in a patient’s glucose level, which is more easily controlled with intravenous insulin, may also have a role in increased mortality. A review of 7,000 medical and surgical intensive care unit (ICU) patients showed that the standard deviation of glucose levels was a better predictor of mortality than the mean glucose level.
Traditional sulfonylureas also close the ATP-sensitive potassium channels.15 This results in the inability of the myocardium to protect against reperfusion injury, and will also result in failure of the ST segment to elevate with a full-thickness MI.16 Due to the lack of ST-segment elevation, patients with diabetes and a full-thickness infarct may be deprived of fibrinolytic therapy, which is not approved for use in the absence of ST-segment elevation. Potentially due to fibrinolytics not being utilized, these patients may sustain more myocardial damage than would have occurred if a fibrinolytic had been utilized. Fortunately, the third-generation sulfonylureas glimepiride and glicizide and the meglitanide nateglinide do not close myocardial ATP-sensitive potassium channels and are therefore safer secretagogs to use in a patient with diabetes.
The fourth risk factor for an MI in type 2 diabetes patients is systolic hypertension. Around 75% of type 2 diabetes patients have hypertension due to the combined effects of hyperinsulinemia and hyperglycemia causing sodium retention and hyperinsulinemia, which stimulates the sympathetic nervous system. Patients with diabetes are exquisitely sensitive to blood pressure lowering. In the Hypertension Optimal Therapy (HOT) study, a ence of 4mm in systolic blood pressure resulted in a 51% decrease in cardiac events.17 In the UKPDS, a 10mm difference in systolic blood pressure non-significantly decreased the incidence of MIs by 21% and significantly reduced strokes and HF by 44 and 56%, respectively. In addition, the lower the systolic blood pressure (down to a minimum of 110mmHg), the lower the incidence of fatal and non-fatal stroke, MIs, and HF. Based on the above and other studies, the target blood pressure for those with diabetes is 130/80, and if proteinuria is in excess of 1g per day, 125/75. To achieve these levels, three to five antihypertensives with different modes of action need to be utilized.
Initial therapy for the diabetic hypertensive should be directed to suppression of the renin–angiotensin system (RAS). Blockade of the sympathetic nervous system (SNS) is obtained with a beta blocker (preferably carvedilol, a vaso-dilating beta blocker that does not increase insulin resistance and through its anti-inflammatory effect reduces oxidative stress and improves endothelial function).18 Beta blockers have clearly been shown to decrease mortality in those with diabetes and ischemic heart disease, and therefore should be part of the armamentarium used to treat diabetic hypertension.19 Small doses of thiazide diuretics, α-blockers, and calcium channel blockers complement the antihypertensive effects of blockers of the RAS and SNS. However, these antihypertensive drugs do not provide the pleotrophic effects of RAS and SNS blockers.
While signifying an increased risk for diabetic nephropathy, the presence of microalbuminuria also reflects the presence of endothelial dysfunction and atherosclerosis, and an increased risk for a cardiovascular event. The glomerulus is effectively an arteriole, and leakage of albumin through the glomerulus into the renal tubule is an indication that lipoproteins are leaking into the sub-endothelial space throughout the vascular tree, leading pathologically to atherosclerosis and clinically to cardiac events. The best example of the prognosis associated with microalbuminuria was shown in the Islington Diabetes Survey, where the presence of microalbuminuria signified an increased frequency of coronary artery and peripheral vascular disease (and after 3.5 years a 16.5-fold increase in mortality) in those with or without diabetes.20 Therefore, in the presence of microalbuminuria, it is mandatory not only to maximize therapy to lower urine albumin, but in addition to maximize therapies for other cardiovascular risk factors. The Steno-2 study was performed in a high-risk microalbuminuric population with type 2 diabetes. In this study, improved blood pressure, glucose, and lipid control resulted in improved cardiovascular and microvascular outcomes (these decreased by 50% after nine years and by 59% after 13.3 years).21
The Detection of Ischemia in Asymptomatic Diabetes (DIAD) study showed that after three years of intensive medical therapy, 79% of diabetes patients with silent ischemia no longer showed evidence of myocardial ischemia with single photon emission computed tomography (SPECT) imaging.22 Therefore, intensive risk reduction can not only prevent (as in the Steno-2 study) but can also reverse (as in the DIAD study) myocardial ischemia. If we make the assumption that everybody with type 2 diabetes has extensive coronary artery disease, we should be confident that aggressive therapy of glycemia, hypertension, and hyperlipidemia will not only prevent the advance of atherosclerosis, but will also stabilize or reverse atherosclerosis and decrease cardiac events. In addition, based on results of the DIAD study, stress testing should not be utilized in asymptomatic diabetic subjects when cardiac risk factors are being aggressively treated.
Increased Platelet Aggregation
Resistance to the antiplatelet effect of aspirin is more common in those with diabetes. For example, a study of 2,500 acute coronary syndrome patients showed that with aspirin there was a significant (48%) reduction in mortality in non-diabetic subjects but no significant reduction in those with diabetes.23 Testing for aspirin resistance is not generally available or utilized, so high-dose aspirin or the addition of another antiplatelet therapy is an option in patients with diabetes. However, unless contraindicated, every patient with type 2 diabetes should be taking at least 81mg of aspirin per day.
Other, Less Traditional Risk Factors in Patients with Type 2 Diabetes
Even with the availability of suppressors of the RAS and SNS, aspirin and statins, and better glycemic, lipid, and hypertension control over the past decade, decreases in mortality are lower in men with diabetes than in men without the disease, and rates have worsened in women with diabetes. There are multiple reasons for this.
Insulin Resistance/Metabolic Syndrome
If the underlying insulin resistance syndrome is untreated in the diabetic patient, this could be responsible for the decreased improvement in mortality.24 The thiazolidinedione pioglitazone has been shown not only to decrease insulin resistance and related cardiac risk factors, but also to reduce the volume of atherosclerosis in both the carotid arteries (the Chicago study) and—as shown on intravascular ultrasound—in the coronary arteries (the Periscope study). In addition, a composite of myocardial infarction, stroke, and death was significantly decreased in the PROactive study of patients with type 2 diabetes and cardiovascular disease when pioglitazone was used.25 Furthermore, in the Prospective Pioglitazone Clinical Trial In Macrovascular Events (PROactive) study, recurrences of myocardial infarction, acute coronary syndrome, and stroke were also reduced.
The other available thiazolidinedione, rosiglitazone, also decreases insulin-resistance-related cardiac risk factors and carotid intima-medial thickening, but does not decrease and may even increase the incidence of myocardial infarction and mortality.26,27 Together, rosiglitazone and pioglitazone (through stimulation of the P-par gamma receptor in the nucleus) activate 23 genes. Five genes are exclusively activated by rosiglitazone, while 12 genes are exclusively activated by pioglitazone. This activation of different genes could explain the differences in the cardiac outcomes of these drugs. Phenotypically, the differences in the lipid profile, which favor pioglitazone, could also be responsible for the differences in cardiac events. Rosiglitazone increases the number of LDL cholesterol particles and apolipoprotein B (ApoB), whereas pioglitazone has the opposite effect. Greater improvements in triglycerides, HDL cholesterol, non-HDL cholesterol, and LDL cholesterol particle size have also been documented with pioglitazone.28
Left Ventricular Hypertrophy
Abnormalities of myocardial structure are not being diagnosed or treated in those with diabetes. Structurally, left ventricular hypertrophy (LVH) is present in 32% of non-hypertensive new diabetes patients and 71% of established diabetes patients.29 The presence of LVH is associated with even greater mortality than multiple vessel coronary artery disease or a decreased ventricular ejection fraction. The increased prevalence of LVH in those with type 2 diabetes is thought to be due to the hyperinsulinemia associated with insulin resistance, since insulin is a growth factor and the prevalence of LVH is increased even in the euglycemic insulin-resistant subject. The addition of uncontrolled hypertension exacerbates the effect of hyperinsulinemia on ventricular growth. The reversal of LVH is achieved with RAS inhibitors, calcium channel blockers, and, to a lesser extent, beta blockers. However, with the reversal of LVH, prognosis improves dramatically.
Ventricular dysfunction is also underdiagnosed in type 2 diabetes patients. Diastolic dysfunction is present in 50–60% of type 2 diabetes patients, and the most likely cause of this ‘stiffening’ of the ventricle is myocardial fibrosis.30,31 The degree of myocardial fibrosis is closely related to chronic hyperglycemia, which leads to glycosylation of proteins, premature crosslinking of collagen, and premature myocardial fibrosis. Myocardial fibrosis is also associated with microalbuminuria. The presence of microalbuminuria suggests endothelial dysfunction where fibrosis results from either extravascular leakage from the microcirculation and/or recurrent microvascular vasoconstriction, leading to microinfarctions and reperfusion injuries with subsequent vasodilation.30
The ‘toxic triad’ of extensive coronary artery disease, LVH, and diastolic dysfunction combine to cause severe ventricular dysfunction, which leads to the activation of the RAS and SNS.32 Activation of the RAS and SNS leads to remodeling of the ventricle from an efficient elliptical shape to an inefficient oval or rotund shape. If ventricular remodeling is not arrested or reversed, pump failure, arrhythmias, and death occur. With these changes, it is not surprising that 43% of patients in the US admitted to hospital with HF have diabetes.33 The reversal of remodeling, and therefore treatment of HF, can best be achieved utilizing RAS inhibitors and beta blockers, which are as effective in patients with diabestes as those without diabetes and HF.34,35
Treatment of the risk factors of hyperglycemia, including post-prandial hyperglycemia, dyslipidemia, and systolic hypertension, has been shown to reduce cardiac events and mortality in patients with type 2 diabetes. In patients with microalbuminuria, therapy should be even more intense since these patients are at higher risk for a cardiac event. Attention to treatment of the insulin resistance syndrome and antiplatelet therapy is also important. Early diagnosis and therapy of LVH and diastolic dysfunction will improve ventricular function and avoid or delay the onset of HF. Early utilization of RAS and SNS inhibitors will not only lower blood pressure but also prevent, suppress, or reverse myocardial remodeling, which in turn will preserve or improve ventricular function and decrease the risk of arrhythmias, HF, and death.■