CHAPTER 28: DIABETES AND CARDIOVASCULAR DISEASE

Cardiovascular disease continues to be a major cause of morbidity and mortality in patients with diabetes and prediabetes, and the prevalence of diabetes and prediabetes is expected to exponentially increase over the next 30 years. The global burden of cardiovascular disease has emerged, not only in the developed world but also in low- and middle-income countries. As a result, the major focus in type 1 diabetes patients and type 2 diabetes patients is the prevention of the development of cardiovascular disease. Recently, the American Diabetes Association (ADA) published their recommendations for practice regarding cardiovascular disease risk management in Diabetes Care in January 2016.¹

The universally recognized criteria for the diagnosis of prediabetes and diabetes are as follows: (1) a normal glycated hemoglobin should be less than 5.7%, (2) a glycated hemoglobin of 5.7% to 6.4% is considered prediabetes, and (3) a level of 6.5% or higher on two separate occasions indicates diabetes.¹ Prediabetes can be diagnosed by a fasting plasma glucose of 100 mg/dL or more (impaired fasting glucose), a postglucose load of 140 to 199 mg/dL (impaired glucose tolerance), or both. Other recognized criteria include a fasting plasma glucose greater than 126 mg/dL or a 2-hour postprandial glucose of greater than 200 mg/dL during an oral glucose tolerance test involving a glucose solution containing the equivalent of 75 grams of glucose dissolved in water. Diabetes can also be diagnosed in the patient who has a random plasma glucose of greater than 200 mg/dL with classic symptoms of hyperglycemia.

With regard to classification and diagnosis of diabetes, a number of important classification issues need to be clarified. Apart from the classification of type 1 and type 2 diabetes, gestational diabetes is the diagnosis of diabetes in the absence of overt diabetes made in the second or third trimester of pregnancy.

The new recommendations also identified patients who are at increased risk for diabetes and who require testing. These include patients who were overweight or obese, with a body mass index (BMI) of greater than 25 kg/m² or, in the case of Asian Americans, of 23 kg/m² or higher. Recommendations for screening for diabetes have now expanded to screening of children and adolescents who are overweight or obese. With regard to gestational diabetes, patients are tested in the first prenatal visit with risk factors. At 28 weeks of gestation, there is a test for gestational diabetes in pregnant women who are not previously known to have diabetes. It is recommended that women with a history of gestational diabetes are considered to have prediabetes and should receive lifestyle interventions for the prevention of diabetes.

Patients with similar mean glucose or glycosylated glycated hemoglobin values can have markedly different daily glucose profiles depending on the numbers and duration of glucose excursions. In 1970, Service and colleagues described the metric Mean Amplitude Glucose Excursions (MAGE) using hourly blood glucose sampling for 48 hours.² It is known that the MAGE calculated from continuous glucose monitoring data corresponds to the originally developed value. A strong relationship between glucose variability and oxidative stress (r = 0.86; P < .001) was observed.³ Furthermore, Marling and coworkers found that the number of excessively variable days using a naive Bayes classifier may be a reliable way to assess the glycemic variability of continuous glucose monitoring (CGM) data.⁴ This measurement closely mirrors 90-day changes in serum 1.5 anhydroglucitrol. Actuation of oxidative stress by acute glucose fluctuations compared with sustained chronic hyperglycemic is observed in patients with type 2 but not in type 1 diabetes.⁵

Globally, diabetes mellitus is a major threat to human health. The number of people with diabetes has increased alarmingly since 1985 and the rate of new cases is escalating. In 1985, an estimated 30 million people worldwide had diabetes and this figure is expected to rise to almost 350 million by 2025.⁶ In the United States in 2012, the unadjusted prevalence of diabetes was 12.3% (95% confidence interval [CI], 10.8%-14.1%). Within the diabetes population, 25.2% (95% CI, 21.1%-29.8%) were undiagnosed.⁷

The prevalence of diabetes is higher in developed countries than in developing countries, but the developing world will be hit the hardest by the diabetes epidemic in the future. Increased urbanization, westernization, economic growth in developing countries have already contributed to a substantial rise in diabetes by promoting overweight, obesity and decreased physical activity. Although diabetes is most common among the elderly in many populations, prevalence rates are rising among young populations in the developing world.

Diabetes mellitus, whether type 1 or type 2, is a very strong risk factor for the development of coronary artery disease (CAD) and stroke⁶ (Table 28–1). Diabetes accelerates the natural course of atherosclerosis in all groups of patients and involves a greater number of coronary vessels with more diffuse atherosclerotic lesions (Fig. 28–1). Cardiac catheterizations in diabetic patients have shown significantly more severe proximal and distal coronary heart disease (CHD) than in patients with CHD but no diabetes. In addition, plaque ulceration and thrombosis have been found to be significantly higher in diabetic patients. Cardiovascular complications include CHD, peripheral artery disease (PAD), nephropathy, retinopathy, cardiomyopathy, and possible neuropathy (involvement of vasa vasorum). These observations underscore the heightened risks of a diabetic patient of developing vascular disease and compel the physician to monitor and correct all risk factors (see Table 28–1). By understanding the mechanisms underlying all these risks, physicians will be poised to prevent them.

The metabolic syndrome is considered to be an aggregation of various established cardiovascular risk factors. Multiple attempts have been made to define the metabolic syndrome encompassing what each investigator deems the relevant clusters. The most widely used definitions come from the National Cholesterol Education Program Adult Treatment Panel (NCEP-ATP) III and the World Health Organization (WHO).^8,9 Other definitions from the International Diabetes Federation and American Association of Clinical Endocrinologists (AACE) have also been proposed.^10,11 Figure 28–1 outlines the relationship between energy excess and obesity and risk factors for the metabolic syndrome.¹²

Microvascular Complications

Renal

Nephropathy occurs in 40% of patients with type 1 and type 2 diabetes. Risk factors include poor glycemic control, hypertension, and ethnicity (blacks, Mexicans, Pima Indians).¹³ Table 28–2 summarizes the key points for the assessment of renal status in a diabetic patient. The earliest clinical finding of diabetic kidney disease is microalbuminuria, which may occur at a time when renal histology is essentially normal (Fig. 28–2). The Diabetes Control and Complications Trial (DCCT) and the United Kingdom Prospective Diabetes Study (UKPDS) showed that the development and progression of microalbuminuria can be prevented through strict glycemic control.

There is insufficient evidence to recommend angiotensin-converting enzyme (ACE) inhibitors in normotensive patients without microalbuminuria. Nonetheless, physicians should still recommend screening on at least a yearly basis, because the risk-to-benefit ratio of diagnosing microalbuminuria justifies treatment with an ACE inhibitor, if not for renal disease alone, then for reducing the incidence of myocardial infarction (MI).

Clinical trials evaluating angiotensin receptor blockers (ARBs), including losartan and irbesartan, have demonstrated a significant renal protective effect in the diabetic patient with nephropathy. There were no differences between the ARB and usual care groups with regard to cardiovascular outcomes.¹⁴

An optimal approach toward diabetic nephropathy combines control of hypertension, preferably with an ACE inhibitor or ARB, glycemic control, sodium restriction, and adjustment of protein intake.

If increasing macroalbuminuria occurs or if renal insufficiency is progressive despite these measures, the patient should be referred to a nephrologist. It is strongly recommended that renal arteriography be avoided. Dietary protein restriction in patients who have progressive renal insufficiency will reduce accumulation of nitrogen-containing waste products and can have a beneficial influence on progression of renal insufficiency.

Renal Hyperfiltration

Glomerular hyperfiltration is the hallmark of early renal hemodynamic disturbance and is an independent predictor of the development and progression of diabetic nephropathy in hypertensive type 2 diabetes when only microalbuminuria is present.¹⁵ Sodium/glucose cotransporter 2 (SGLT-2) inhibitors appear to delay progression to albuminuria in type 2 diabetic patients with chronic kidney disease.^16,17,18,19

Ophthalmologic

Diabetic retinopathy is the most prevalent microvascular complication, affecting nearly 50% of the diabetic population at a given time and eventually occurring in all diabetic patients. Diabetes remains the leading cause of visual loss in adults. Visual loss from diabetes occurs either as a result of proliferative retinopathy or macular edema.²⁰

Macrovascular Manifestations

Coronary Heart Disease

CHD is strongly associated with type 2 diabetes mellitus and is the leading cause of death regardless of the duration of disease. There is a two- to fourfold increase in the relative risk ratio of cardiovascular disease in type 2 diabetes patients compared to the general population. This increase is particularly disproportionate in diabetic women. The protection that premenopausal women have against CHD is not seen if they suffer from diabetes. The degree of hyperglycemia and the duration of hyperglycemia are strong risk factors for the development of microvascular and macrovascular complications. Even impaired glucose tolerance increases cardiovascular risk, although there is minimal hyperglycemia.²¹

The first detectable sign of a problem in people genetically prone to develop type 2 diabetes is insulin resistance, which can be seen as long as 15 to 25 years before the onset of diabetes. Several atherogenic factors are associated with insulin resistance, which can start the atherosclerotic process years before clinical hyperglycemia ensues. It is unclear whether the compensatory hyperinsulinemia plays a role in atherosclerosis generation in insulin-resistant patients. A number of prospective studies have shown an association between fasting or postprandial hyperinsulinemia and the future development of CHD.

Hyperglycemia itself plays an important role in enhancing the progression of atherosclerosis in type 2 diabetes. The threshold above which hyperglycemia becomes atherogenic is not known but may be in the range defined as impaired glucose tolerance (ie, fasting plasma glucose level < 126 mg/dL with 30-, 60-, or 90-minute plasma glucose concentrations > 200 mg/L and a 2-hour plasma glucose level of 140 to 200 mg/dL during an oral glucose tolerance test). An important Mendelian randomization study strongly suggests that not only is type 2 diabetes mellitus directly causal for CHD but that longer term studies are needed to demonstrate a benefit of glucose lowering on CHD outcomes.²²

Acute Coronary Syndromes

Diabetic patients represent a high-risk group for developing and surviving acute MI. In particular, patients with type 1 diabetes have a worse outcome than do patients with type 2 disease, and diabetic women have almost twice the risk of death of diabetic men. Reperfusion therapy is the cornerstone of the management of acute MI. In a meta-analysis of the major trials comparing thrombolytic therapy to percutaneous coronary intervention (PCI), diabetic patients had significantly higher 30-day mortality when compared with nondiabetic patients (9.4% vs 5.9%; P < .001).²³ The advantages of angioplasty over thrombolytic therapy in the diabetic population are comparable to those observed in the nondiabetic patients.

New treatment strategies are emerging. The use of insulin and glucose infusion for at least 24 hours after admission followed by intensive long-term insulin was compared with usual care in the Diabetes Mellitus Insulin-Glucose in Acute Myocardial Infarction (DIGAMI) trial. A total of 620 diabetic patients were randomized, and the trial demonstrated a 30% reduction in mortality at 12 months for the group treated under the intensive program.²⁴

However, the follow-up DIGAMI-2 trial, which compared three strategies—intensive insulin infusion followed by long-term insulin therapy, intensive insulin infusion followed by long-term standard glucose control, and regular metabolic control—in 1253 patients was unable to show a difference in short- and long-term mortality and morbidity between any of the arms. DIGAMI-2 concluded that intensive glucose control was of great importance in the peri-ischemic period irrespective of the strategy used to achieve it.²⁵

The use of antiplatelet therapy is the mainstay of management of acute coronary syndromes in diabetic patients. With the advent of newer antiplatelet drugs, there have been recommendations for greater utilization of prasugrel and ticagrelor. The evidence suggests that there is no heterogeneity in the response to newer agents and strategies based on diabetes status^{26,27,28,29,30,31} (Table 28–3³²). The optimal strategy and regimen in diabetic patients is still elusive and remains the question of ongoing trials focused on limiting the thrombotic burden and not increasing the risk of major bleeding.

Chronic Coronary Artery Disease

The association between CHD and diabetes is strong and has led to screening strategies in diabetic patients even before they are symptomatic. In addition, diabetic patients often are unaware of myocardial ischemic pain, and so silent MI and ischemia are markedly increased in this population. There is a heightened concern for the development of sudden cardiac death in those with diabetes. Therapeutic modalities in diabetic patients with CHD revolve around standard therapy with aspirin, β-blockers, calcium channel blockers, and nitrates.

Epidemiologic evidence from the Bezafibrate Infarction Prevention Study registry shows almost a 50% reduction in mortality for type 2 patients with chronic CHD who were treated with β-blockers, compared with controls. Other randomized trial evidence has demonstrated that diabetes is a strong predictor of death and that diabetic patients may benefit more from β-blocker therapy than do nondiabetics. In general, β-blockers are extremely well tolerated, and masking or prolonging of hypoglycemic symptoms appears to be highly infrequent, particularly with cardioselective β blockers.³³

Novel therapies have been tested to address the increased ischemic burden experienced by diabetic patients with CAD. Ranolazine has been studied widely and has recently been shown to lower the weekly angina frequency when compared to placebo (3.8 vs 4.3 episodes per week; P = .008) as well as the use of sublingual nitroglycerin in the Type 2 diabetes Evaluation of Ranolazine in Subjects with Chronic Stable Angina (TERISA) trial.³⁴

Diabetes-Related Cardiomyopathy

Diabetes-related cardiomyopathy is a term used by clinicians to encompass the multifactorial etiologies of diabetes-related left ventricular failure characterized by both systolic and diastolic function.³⁵ The Framingham Heart Study showed that men with diabetes who have congestive heart failure were twice as common as their nondiabetic counterpart and that females with diabetes had a fivefold increase in the rate of congestive heart failure. The spectrum of heart failure ranges from asymptomatic to overt systolic failure. Diabetes complicated by hypertension represents a particularly high-risk group for the development of congestive heart failure.³⁶ Diastolic dysfunction is exceedingly common (> 50% prevalence in some studies) and may be linked to diabetes without the presence of concomitant hypertension.

The etiology of impaired left ventricular function may involve any of the following mechanisms: (1) coronary atherosclerotic disease, (2) hypertension, (3) left ventricular hypertrophy, (4) obesity, (5) endothelial dysfunction, (6) coronary microvasculature disease, (7) autonomic dysfunction, and (8) metabolic abnormalities.

Echocardiographic studies confirm that diastolic abnormalities occur in young diabetic patients who have no known diabetic complications.³⁷ Presently diabetic cardiomyopathy is included under the umbrella of heart failure with preserved left ventricular ejection fraction (HFpEF), which is characterized by coronary microvascular endothelial dysfunction related to hyperglycemia, hyperinsulinemia, and lipotoxicity and leading to a restrictive picture.³⁸ Patients will have impaired diastolic dysfunction, and valvular, coronary, and hypertensive causes should be excluded. The management of HFpEF is diuretic therapy and usually includes β-blockers and ACE inhibitors, but evidence for this is sparse.

Once CHD develops in the diabetic patient, the heart failure with preserved left ventricular systolic dysfunction (HFpEF) that ensues responds to all the same therapies as in the nondiabetic population. The findings of the Heart Outcomes Prevention Evaluation (HOPE) trial suggest that early initiation of ACE inhibition retards the progression to overt congestive heart failure.³⁹

Cerebrovascular Disease

Compared to nondiabetic subjects, the mortality from stroke in diabetic patients is almost threefold higher. The small paramedial penetrating arteries are the most common sites of cerebrovascular disease. In addition, diabetes increases the likelihood of severe carotid atherosclerosis. Diabetic patients are likely to suffer increased brain damage with carotid emboli that would result in a transient ischemic attack in a nondiabetic individual.⁴⁰

The spectrum of metabolic disturbances associated with diabetes and insulin resistance extends beyond hyperglycemia and includes dyslipidemia, hypercoagulability, and inflammation.

Adipokines

Adipose tissue is an active organ that produces multiple adipokines,⁴¹ which play an important role in the development and progression of atherosclerosis. Table 28–4 shows the prominent adipokines with their actions.

Endothelial Dysfunction

The normal endothelium regulates vasomotor tone, and keeps the coagulation cascade in balance. Insulin resistance, itself, appears to be an endothelial dysfunction risk equivalent.

In healthy endothelium, the stimulation of insulin receptors activates the phosphoinositol-3 kinase (PI3-K) pathway. The production of nitric oxide by endothelial cells stimulated by this pathway leads to a host of anti-inflammatory and antithrombotic effects that are antiatherogenic. The endothelial dysfunction of insulin resistance is a result of decreased levels of nitric oxide and thus impaired blood flow because of a downregulation of the PI3-K pathway.⁴²

A parallel pathway downregulates the production of endothelin-1. In response to hyperglycemia, endothelial cells secrete endothelin-1, which induces proatherogenic effects such as vasoconstriction, increased vascular permeability, and vascular smooth muscle cell (VSMC) proliferation. Elevated glucose levels stimulate the production of interleukin-6 by endothelial cells and monocytes and increased proteoglycan synthesis by VSMCs.

Vascular Smooth Muscle Cell Migration and Proliferation

The insulin receptors on VSMCs are structurally and functionally similar to those in skeletal muscle and adipocytes. Insulin signaling in VSMCs initiates proatherogenic cellular events such as proliferation and migration. Insulin acts synergistically with other atherogenic growth factors, such as platelet-derived growth factor, to promote the proliferation and migration of bovine aortic smooth muscle cells.

Monocyte/Macrophage Adhesion and Migration

Circulating monocytes are important inflammatory cells involved in the immune response. Insulin and insulin-like growth factor-1 receptors are present on circulating monocytes and macrophages. Defective insulin signaling is implicated in macrophage foam cell formation.⁴³ Under conditions of insulin resistance, the protective effect of insulin to reduce macrophage apoptosis may be lost. Macrophage apoptosis in atherosclerotic lesions may contribute to further monocyte recruitment by the release of cytokines and may thus accelerate the development of the vascular lesion and contribute to plaque rupture.

Insulin Resistance

The insulin resistance syndrome is a composite of dyslipidemia, hypertension, and hypercoagulability.⁴⁴ It is only now being recognized that insulin resistance is the predominant defect in more than 90% of type 2 diabetes patients and the major pathologic mechanism for the susceptibility to premature cardiovascular disease. Insulin resistance and hyperinsulinemia accelerate the development of atherosclerosis. Hyperinsulinemia is an independent risk factor when adjusted for lipid profile, hypertension, and family history. Studies of multiple ethnic groups show increased carotid intima-medial thickness (a reliable marker for coronary disease) in subjects with insulin resistance. Impaired glucose tolerance can increase the risk of heart disease. Because insulin resistance precedes clinically diagnosed type 2 diabetes by 10 to 15 years in as many as 90% of patients, this extensive period of atherogenic exposure may account for the higher rates of cardiovascular disease in type 2 diabetics.^45,46

Although the euglycemic-insulin clamp remains the gold standard for the measurement of insulin resistance, simpler formulas are used clinically to quantify insulin resistance, or IR, using the Homeostasis Model Assessment (HOMA IR = Serum Insulin × Serum Glucose/22.5).⁴⁵ Insulin resistance as assessed by HOMA (HOMA-IR) also has been shown to be predictive of cardiovascular disease. In long-term follow-up of type 2 diabetic patients, insulin resistance was independently predictive of cardiovascular disease, with a one-unit increase in HOMA-IR associated with a 5.4% increased risk of cardiovascular disease.⁴⁵ The Insulin Resistance and Atherosclerosis Study (IRAS) also demonstrated the relationship between insulin resistance and atherosclerosis in the carotid artery.

Diet and Lifestyle Modification

There have been major advances in our understanding of the role of lifestyle modification in diabetic patients.

Benefits of Modest Weight Loss

The current approach to managing type 2 diabetes is to focus on modest weight loss in the range of 5% to 10% of initial weight. The concomitant reductions in low-density lipoprotein (LDL) cholesterol and systolic blood pressure along with improvements in glycated hemoglobin have real and sustained cardiovascular benefits.⁴⁷ Weight loss is an important therapeutic strategy in all overweight or obese individuals who have type 2 diabetes or are at risk for developing diabetes.⁴⁸ The primary approach for achieving weight loss, in the vast majority of cases, is therapeutic lifestyle change, which includes a reduction in energy intake and an increase in physical activity. A moderate decrease in caloric balance (500-1000 kcal/d) will result in a slow but progressive weight loss (1-2 lb/wk). For most patients, weight-loss diets should supply at least 1000 to 1200 kcal/d for women and 1200 to 1600 kcal/d for men. In selected patients, drug therapy to achieve weight loss as an adjunct to lifestyle change may be appropriate. However, it is important to note that regain of weight commonly occurs on discontinuation of medication.

Physical activity is an important component of a comprehensive weight management program. Regular moderate-intensity physical activity enhances long-term weight maintenance. Regular activity also improves insulin sensitivity, glycemic control, and selected risk factors for cardiovascular disease (ie, hypertension and dyslipidemia), and increased aerobic fitness decreases the risk of CHD. Initial physical activity recommendations should be modest, based on the patient’s willingness and ability, gradually increasing the duration and frequency to 30 to 45 minutes of moderate aerobic activity, 3 to 5 days per week, when possible. Greater activity levels of at least 1 hour per day of moderate (walking) or 30 minutes per day of vigorous (jogging) activity may be needed to achieve successful long-term weight loss. The American College of Sports Medicine now recommends that resistance training be included in fitness programs for adults with type 2 diabetes.⁴⁹ Resistance exercise improves insulin sensitivity to about the same extent as aerobic exercise.

Look-AHEAD

The National Institutes of Health–sponsored The Action for Health in Diabetics (Look-AHEAD) trial of weight reduction and exercise in over 5000 type 2 diabetic patients sought to evaluate the role of modest weight loss and an exercise program on major cardiovascular events with at least 8 years of follow-up (Fig. 28–3).⁵⁰ The trial was halted prematurely in September 2012 because of futility, bringing into question whether lifestyle provides adequate cardioprotection in type 2 diabetes. The surrogate markers appeared favorable with a greater rate of modest weight loss in the intervention arm (4.7% vs 2.1% of initial weight was lost by year 8; P < .001) and with more than one-fourth of patients in the intervention group losing at least 10% of initial weight in the long-term. The real challenge in interpreting Look-AHEAD is that the control group was aggressively treated for cardiovascular risk through marked cointervention of evidence-based therapies.

PREDIMED

The Mediterranean diet, originally coined by Ansel Keys, is rich in fruits, vegetables, and olive oil, with the addition of chick peas, bulgur, couscous, lentils, and fava beans. It is very limited in the consumption of red meat and eggs. Emerging evidence has shown that adoption of the Mediterranean diet has positive effects on the progression of type 2 diabetes and on cardiovascular outcomes. Shia and colleagues first showed that rates of type 2 diabetes development and progression can be markedly reduced.⁵¹ The transformational Prevención con Dieta Mediterránea (PREDIMED) study in Spain demonstrated a 52% reduction (95% CI, 0.27-0.86) in the development of type 2 diabetes mellitus in the primary prevention cohort⁵² (Fig. 28–4). At present, the evidence for the preventive effects of the Mediterranean diet in type 2 diabetes mellitus is rather limited. A case-control Italian study of 144 type 2 diabetes mellitus patients with PAD matched to 288 age- and sex-matched type 2 diabetes mellitus controls showed that a diet with a greater adherence to Mediterranean diet was associated with a lower (by 50%) incidence of PAD.⁵³

STENO-2 demonstrated that a comprehensive multifactorial strategy (including lifestyle and pharmacologic interventions) to reduce cardiovascular risk in type 2 diabetic patients with microalbuminuria was highly effective (hazard ratio [HR] 0.47; 95% CI, 0.24-0.73) when compared to usual care after a mean time of 7.8 years. The number needed to treat to prevent a major cardiovascular event was only five patients. The approach included targets of glycated hemoglobin less than 6.5%, blood pressure less than 130/80 mm Hg, total cholesterol less than 175 mg/dL, and triglycerides below 150 mg/dL. Patients were prescribed aspirin and an ACE inhibitor or ARB. This study validates the multidisciplinary approach to the cardiovascular care of the diabetic patient.⁵⁴

Dysglycemia

Glycemic Control

The ADA has increased the recommendation for premeal blood glucose targets to 80 to 130 mm/dL (up from 70 to 130 mm/dL). The ADA has also promoted the utilization of CGM into practice and laid out a plan to educate on CGM.¹

Prevention or the Delay of Type 2 Diabetes

Patients with an impaired glucose tolerance test or glycated hemoglobin of 5.7 to 6.4% should be recommended for intensive diet and physical activity intervention with a target loss of body weight of about 7%. Metformin should be considered for those with a BMI of more than 35 mg/m² who are under the age of 60 years or in women with prior gestational diabetes. Recommended lifestyle intervention is for 7% weight loss with diet and moderate intensity physical activity of at least 2.5 hours per week.

Dysglycemia Management

Patients with type 1 diabetes (generally caused by viral illness or autoimmunity) are treated with insulin. A cardiologist should know that such a patient will require an average of 0.5 units of insulin/kg body weight per day (half of this is basal insulin [0.25 units/kg] and the other half is bolus insulin (0.25 units/kg), if divided in three boluses injections given before each meal. The patient will need to be managed by an endocrinologist who can work collaboratively with the cardiologist because cardiovascular disease and diabetes are inextricably intertwined.

Glycemic control in type 1 diabetics is associated not only with lower rates of microvascular complications than observed in type 2 diabetes but also with reductions in cardiovascular events. One of the major distinctions between type 1 and type 2 diabetes is the timing of initiation of insulin therapy and the types of insulin therapy that are provided. Patients with type 1 diabetes are prescribed a balanced individualized diet and exercise program and are often treated with multiple-dose insulin (known as MDI injections), which include the use the insulin analogs.

Patients with type 2 prediabetes or diabetes are managed by first teaching them lifestyle changes, including a Mediterranean diet and carbohydrate counting with an aerobic exercise program at last four times a week. The training is shared between an endocrinologist and a certified diabetes nurse or nutritionist educator, either of whom should be a certified diabetes educator. The target glycated hemoglobin is generally 6.5% (AACE) or less than 7% (ADA, European Association for Study of Diabetes [EASD]) unless higher as dictated by a personalized patient-centered approach, which is discussed. In some cases, glycated hemoglobin of 8.5% is acceptable with a patient who is elderly, has multiple comorbidities, limited resources, and an anticipated short life span. We offer two algorithms for the management of type 2 diabetes from ADA (Fig. 28–5) and from AACE (Fig. 28–6). The AACE algorithm has gained more traction as it is more pragmatic. Metformin is generally the first drug to be used, with add-ons depending on the circumstances of each patient. Basal insulin and eventually basal and bolus insulin before meals is generally required in most, but not all, patients. Mindful eating and lifestyle changes are of critical importance. It is best to refer most patients to an endocrinologist for consultation because the algorithm can be daunting to a busy cardiologist or internist, and a nuanced personalized approach is strongly recommended.¹

More recently, Inzucchi and colleagues have reviewed their valuable patient-centered approach to the management of type 2 diabetes.⁵⁵ With the ADA guideline for a target glycated hemoglobin of 7%, factors that would lead a clinician to aim to a higher target in decision making were the following: risks potentially associated with hypoglycemia and other drug effects, disease duration, life expectancy, important comorbidities, established vascular complications, patient attitude and expected treatment effects, and resources and support system (Fig. 28–7). There are some patients with multiple such factors in whom a glycated hemoglobin of 8.5% might be reasonable to prevent untoward outcomes.

Hypoglycemia

One of the most important aspects of glycemic control is not only to prevent microvascular events but also to avoid hypoglycemia. Once a hypoglycemic episode has occurred, patients should be reevaluated for their treatment strategy. Patients who are elderly should have ongoing cognitive assessment to make sure that they are aware of hypoglycemia. Hypoglycemia is associated with a greater risk of dementia and cognitive impairment.⁵⁶ In the Action to Control Cardiovascular Risk in Diabetes Study Group (ACCORD) and The Action in Diabetes and Vascular Disease: Preterax and Diamicron MR Controlled Evaluation (ADVANCE) trials, severe hypoglycemia was associated with increased overall mortality.^57,58

Antidiabetic Medications

The Use of Insulin

In patients with type 2 diabetes, the use of insulin is largely related to the duration of type 2 diabetes. As patients survive longer with diabetes given the optimization of medical therapy, there is a greater likelihood that patients will eventually require insulin in order to maintain their target glycated hemoglobin. It is important to remember that type 2 diabetes is a progressive disease and that eventually patients will require more intensive medical therapy, including insulin.

Insulin is used in the management of type 1 diabetes mellitus or type 2 diabetes mellitus as monotherapy or in combination with oral agents. Insulins currently available are synthetic human insulins or analogs of human insulin, which differ in their rate of absorption and duration of action. There are also products that are mixtures of rapid short-acting and intermediate-acting insulins (Table 28–5). Purified animal insulins are no longer used.

Basal insulin can be initiated at 10 units per day or 0.1 to 0.2 units per kg/d. Then basal insulin is adjusted 10% to 15% or two or four units once or twice weekly to reach the fasting blood glucose target. Although this is not the topic of this chapter, it is important to note that working closely with the endocrinologists will allow cardiovascular specialists to be more involved in the care of their patients, including those for whom insulin is prescribed. When more complex insulin regimens are used beyond basal insulin strategies, combination of insulin with oral antidiabetic drugs is common.

Use of Insulin Pumps and Continuous Glucose Monitoring Systems

Insulin pumps are devices with a subcutaneous catheter that deliver continuous subcutaneous insulin infusion. One or more basal rates are preprogrammed by the user, and boluses are taken as needed whenever carbohydrates are ingested. The catheter is changed every 2 to 3 days, and abdominal infusion sites are most commonly used. In a motivated patient, better glycemic control can be achieved with continuous subcutaneous insulin infusion—compared with multiple subcutaneous insulin injections—because continuous subcutaneous insulin infusion can provide multiple basal rates of insulin.⁵⁹

CGM systems that use a glucose sensor to provide up to 3 days of CGM in the subcutaneous tissue are available. The record shows glucose patterns and trends that can help in the recognition and prevention of hypoglycemia, hyperglycemia, postprandial glucose excursions, nocturnal hypoglycemia, and the effects of exercise. However, “normal” interstitial glucose values may be lower than realized, including some values in the “hypoglycemic range.” Also, the CGM systems may not always read the glucose concentrations consistently and accurately. A recent comprehensive consensus document reports on the use of CGM systems in type 1 and type 2 diabetes. More trials are needed in type 2 diabetes. Insulin delivery can be integrated with digital technology to change insulin dosing and avoid hypoglycemia.

Multiple other agents are used in the management of diabetes. Table 28–6 outlines the various oral agents used to control glycemia in diabetes. The ominous octet described by Defronzo dictates that multiple agents tackling the various pathophysiological defects are necessary to managing type 2 diabetes⁶⁰ (Fig. 28–8).

Metformin

Metformin⁶¹ has been used in Europe for several decades but has been marketed in the United States since 1995. The mechanism of action of metformin is to decrease hepatic glucose output by inhibiting glucose-6-dehydrogenase activity and stimulating the insulin-induced component of glucose uptake into skeletal muscle and adipocytes and decrease the intestinal absorption of glucose.

The starting dose for metformin is 500 mg orally with dinner. A sustained-release preparation is preferred for once-daily dosing and should be uptitrated weekly to two, three, and four 500-mg tablets at dinner. Because of its mechanism of action, there is minimal risk of hypoglycemia. This drug should be used with caution and at reduced dosing in renal disease, and not at all in renal failure or potential hypoxic states, such as congestive heart failure and severe pulmonary disease, because of the risk of lactic acidosis.

Thiazolidinediones

Thiazolidinediones (TZDs)^62,63 induce peroxisome proliferator-activated receptor (PPAR)-γ binding to nuclear receptors in muscle and adipocytes, allowing insulin-stimulated glucose transport. Three PPARs have been identified to date: PPARα, PPARδ (also known as PPARβ), and PPARγ. After ligand binding, PPARs change their conformation to permit the recruitment of one or more coactivator proteins. The first mechanism, transactivation, is DNA-dependent and consists of binding PPAR components with target genes and heterodimerization with the retinoid X receptor. The second mechanism, transrepression, interferes with other transcription factor pathways that are not DNA-dependent. The protein kinase C signaling pathway functions as a molecular switch that dissociates with transactivation and transrepression properties of PPARα. PPARα resides mainly in the liver, heart muscle, and vascular endothelium; when it is activated, it controls genes that regulate lipoprotein levels and confers anti-inflammatory effects. PPARγ is located mainly on adipocytes, but it is also found in pancreatic beta cells, vascular endothelium, and macrophages.

Thiazolidinediones and Cardiovascular Outcomes

The TZDs have come under increased scrutiny because of the association with edema and weight gain. What is clear after a number of key trials is that there are likely differences between the two approved PPAR agents, pioglitazone and rosiglitazone.⁶⁴ Pioglitazone and rosiglitazone have significantly different effects on plasma lipids.^65,66 Compared with rosiglitazone, pioglitazone is associated with significant improvements in LDL particle concentration, LDL particle size, triglycerides, and high-density lipoprotein (HDL) cholesterol. When it comes to the TZDs, it is essential to balance the risk and benefits. Clearly, patients with chronic heart failure are not candidates for TZD therapy.

The pivotal outcomes trial of pioglitazone, the Prospective Pioglitazone Clinical Trial in Macrovascular Events (PROactive), did not provide as clear a picture as the surrogate outcomes trials.⁶⁷ This large trial of over 5000 diabetic patients with established macrovascular complications were randomized to pioglitazone 45 mg/d or placebo. The primary end point, which included MACE but also leg amputation, peripheral and coronary revascularization, and acute coronary syndromes, did not demonstrate any difference. However, a key secondary end point of death, MI, and stroke showed a 16% relative risk reduction in favor of pioglitazone (HR 0.84; 95% CI, 0.72-0.98; P = .027). The potential reduction in cardiovascular events associated with pioglitazone is supported by a large randomized, placebo-controlled trial of pioglitazone in patients after stroke or transient ischemic attack. With 4.8 years of follow-up, the rate of MI or stroke was significantly reduced in the pioglitazone arm (9.0% versus 11.8%; HR 0.76; 95% CI, 0.62-0.93).⁶⁸

The evidence for a cardioprotective effect of rosiglitazone is not apparent. In fact, the meta-analysis by Nissen and Wolski created a media firestorm by reporting a 4% excess in MI and a 60% excess for cardiovascular death for patients treated with rosiglitazone compared with an active comparator in type 2 diabetes mellitus.⁶⁹ This meta-analysis was highly criticized for a number of reasons, including pooling trials for which cardiovascular end points were relatively rare, with a great deal of heterogeneity between trials. What was most controversial is that it led to the publishing of an interim analysis of the Rosiglitazone Evaluated for Cardiac Outcome and Regulation of glycemia in Diabetes (RECORD) trial.⁷⁰ Eventually, RECORD was published.⁷¹ The trial randomized over 4000 patients in an open-label design. After a mean follow-up of 5.5 years, the addition of rosiglitazone to a combination of metformin and sulfonylurea was found to be noninferior to metformin and sulfonylurea without rosiglitazone for the combination of cardiovascular hospitalization or cardiovascular death (HR 0.99; 95% CI, 0.85-1.16). Heart failure causing admission to hospital or death was increased in the rosiglitazone group (HR 2.10; 95% CI, 1.35-3.27). At present, rosiglitazone is rarely used.

Sulfonylureas

Sulfonylureas⁷² are the oldest class of treatment for type 2 diabetes mellitus. The mode of action is by stimulating β-cell insulin secretion. The beta-cell sulfonylurea receptor (SUR) is functionally linked to an ATP-sensitive K⁺ channel (K⁺-ATP) on the cell membrane. In the basal state, the K⁺-ATP channel shifts K⁺ from the inside of the β cell to the extracellular space and maintains the resting potential of the beta-cell membrane. When the sulfonylurea binds to the SUR, K⁺ efflux diminishes and the membrane depolarizes. This depolarization opens a voltage-dependent calcium channel in the same membrane, which enables extracellular calcium to enter the cell. The resultant increase in intracellular calcium triggers insulin-containing secretory granule exocytosis.

The first-generation sulfonylureas have a long half-life and bind ionically to plasma proteins, making them easily displaced. The major concern with these agents is hypoglycemia. The second-generation sulfonylureas have a shorter half-life and bind to plasma proteins nonionically, making them less easily displaced from proteins and available for binding to receptors. Commercially available second-generation sulfonylureas are glyburide (1.25-20 mg/d), glipizide (2.5-40 mg/d), and glimepiride (1-8 mg/d). Sulfonylureas decrease the glycated hemoglobin by 1% to 2%.

Incretins

The newer agents available for the treatment of type 2 diabetes mellitus belong to the class of incretin hormones⁷³ (see Table 28–6). The two key incretins are GLP-1 and glucose-dependent insulinotropic peptide (GIP), both secreted in the small intestine by the L cells and K cells, respectively. Both GLP-1 and GIP are decreased in type 2 diabetes mellitus, and when given pharmacologically to animals, these hormones stimulate beta-cell proliferation and can prevent or delay the onset of diabetes.

Exenatide is a synthetic peptide that is a GLP-1 agonist (incretin mimetic). It potentiates insulin secretion and decreases glucagon secretion postprandially. Exenatide delays gastric emptying and promotes satiety, resulting in weight loss. Liraglutide, another in this class, has also been approved as an antiobesity drug by demonstration of over 9% weight loss in the SCALE trial in nondiabetic patients.⁷⁴

Dipeptidyl Peptidase 4 Inhibitors

The enzyme dipeptidyl peptidase 4 (DPP-4) rapidly degrades GLP-1 and GIP to inactive forms and include the following agents: sitagliptin, vildagliptin, saxagliptin, linagliptin, and alogliptin. DPP-4 inhibitors retard peptide degradation of these incretins, allowing therapeutic efficacy.⁷⁵ The DPP-4 inhibitors are given orally and are weight neutral.⁷⁶

In head-to-head comparisons, exenatide has been shown to reduce weight and postprandial triglyceride levels when compared to a DPP-4 inhibitor.⁷⁷

Cardiovascular Benefits of GLP-1 and DPP-4 Inhibitor Agents

For the GLP-1 agonists, weight loss may be a key to improved cardiovascular effects. Indirect benefits can arise through reductions in blood pressure and inflammatory markers. There are GLP-1 receptors in the cardiac myocytes, and treatment with GLP-1 has been shown to improve myocardial function in rat and dog heart failure models.^78,79,80,81

DPP-4 compounds are not associated with either weight gain or reduction and hold promise by enhancing meal time levels of GLP-1 and GIP.⁸² The risk of hypoglycemia is very low. Because hypoglycemia has been linked to acute cardiovascular events, the DPP-4 inhibitors may provide an additional margin for safety.

Over the past 2 years, a number of important studies of a newer class of drugs, the incretin therapies, have emerged. This includes three important studies of a DPP-4 inhibitors and one for a GLP agonist in the treatment of patients with type 2 diabetes at cardiovascular risk. Overall, these trials of incretin drugs that are currently reported have shown a neutral effect on the primary end point of cardiovascular mortality, nonfatal MI, and nonfatal stroke.

The SAVOR TIMI-53 trial (Saxagliptin Assessment of Vascular Outcomes Recorded in Patients with Diabetes Mellitus [SAVOR Trial]–Thrombolysis in Myocardial Infarction [TIMI] 53 trial), evaluating saxagliptin, showed a hazard ratio of 1.0 for primary cardiovascular events out to about 2 years.⁸³ The EXAMINE (EXamination of CArdiovascular OutcoMes: AlogliptIN vs. Standard of CarE in Patients with Type 2 Diabetes Mellitus and Acute Coronary Syndrome) trial, with alogliptin in acute coronary syndrome patients, showed an HR of 0.96 with a 95% CI that encompassed unity.⁸⁴ The TECOS (Trial Evaluating Cardiovascular Outcomes With Sitagliptin) trial, presented at the ADA in June 2015, evaluated the use of sitagliptin and showed an HR of 0.98.⁸⁵

A GLP-1 agonist has also now been tested in a trial. The ELIXA (Evaluation of Lixisenatide in Acute Coronary Syndrome) trial with lixisenatide showed an HR of 1.02 for primary cardiovascular events (95% CI, 0.89-1.17).⁸⁶

In light of the SAVOR trial finding of about a 25% increase in heart failure hospitalization with saxagliptin over placebo, a cardiovascular safety concern has been considered. Recent evidence suggests that this signal was only apparent in the SAVOR trial with the other three outcome trials showing no significant increase in heart failure.⁸⁷ This issue is not fully resolved at present, and recent studies suggest that the link between heart failure and antidiabetic drugs may be tightly linked to weight gain.^88,89,90

SGLT Inhibitors

The SGLT-2 inhibitors block renal glucose resorption in the proximal tubule, thereby increasing urinary glucose excretion. Cardiovascular end points have now been reported for one drug in this class, empagliflozin, which is associated with weight loss and significant reductions in glycated hemoglobin comparable to other oral antihyperglycemic agents, as well as reductions in systolic blood pressure.⁹¹

The EMPA-REG trial was conducted to examine the long-term effects of empagliflozin versus standard of care on cardiovascular morbidity and mortality in type 2 diabetes.⁹² This was a randomized, double-blind, placebo-controlled trial that was conducted at 590 sites in 42 countries. Patients were screened for the trial, and 7020 patients were randomized either to placebo or two doses of empagliflozin 10 mg daily or 25 mg daily in a 1:1:1:1 allocation. At least 691 patients would experience a primary adjudicated event of cardiovascular death, nonfatal MI, and nonfatal stroke. Patients were included if they had a BMI of less than 45 kg/m² and a glycated hemoglobin of 7% to 10% with a history of established cardiovascular disease. Patients were excluded if their estimated glomerular filtration rate was less than 30 mm/m. The primary end point was a three-point MACE of cardiovascular death, nonfatal MI, and nonfatal stroke. The trial recruited about 72% males with an average glycated hemoglobin of 8.1%. About one-half of the patients who were on insulin therapy and the mean BMI in the trial was about 30.7. Only 10% of the patients had a history of cardiac failure prior to enrollment in the study.

The glycated hemoglobin in the trial was not statistically different between the treatment and placebo groups after long-term follow-up of about 4 years. Patients in the first 12 weeks of the study had to stay on their current therapy, so those randomized to placebo had a slightly elevated glycated hemoglobin but by the end of the trial patients were treated to a target glycated hemoglobin of about 7%. The overall mean glycated hemoglobin achieved was closer to 8%. With regard to other important biomarkers, empagliflozin was associated with no difference in heart rate, diastolic blood pressure, or LDL cholesterol. At a median follow-up of 2.6 years, empagliflozin was associated with a reduction in systolic blood pressure of approximately 4 mm Hg, a reduction in waist circumference of approximately of 2 cm, and a reduction in weight of 2 kg. For the primary end point of the study, the HR overall was approximately 0.85 with no significant differences overall between the two doses of empagliflozin. When comparing placebo to empagliflozin overall, there is about 15% reduction in primary events, which just met statistical significance at P value of .04 (see Fig. 28–3). However, there were significant reductions in cardiovascular death (HR 0.62; 95% CI, 0.49-0.77; P < .0001) and a significant reduction in the hospitalization for heart failure (HR 0.065; 95% CI, 0.50-0.85; P = .0017). The drug was safe and well tolerated, and the only significant safety issue was a threefold increase in the rates of genital infections. This was observed in both males and females. The number of confirmed cases of hypoglycemia was no different, at about 28% in both the placebo and treatment arms. The number of patients experiencing diabetic ketoacidosis was not significantly different, at about 0.1% overall.

For patients treated with empagliflozin, the number needed to treat it at 3 years to prevent one death is approximately 39 patients. This compared very similarly to the number needed to treat for simvastatin to prevent one death in the Scandinavian Simvastatin Survival Study (4S) after 5 years of treatment.⁹² The EMPA-REG study will likely transform the care of patients with type 2 diabetes and will lead empagliflozin to now be the first- or second-line agent in diabetic patients with established cardiovascular disease. Table 28–2 summarizes the different classes of drugs used in the treatment of type 2 diabetes. This, combined with Fig. 28–10, provides an algorithm that is approved by the ADA. What is clear in the light of EMPA-REG and the relative safety of the DPP-4 inhibitors is that sulfonylureas are likely to be relegated to fourth or fifth line use.

Results of Large Cardiovascular Outcome Trials Evaluating the Role of Glucose Control

For type 1 diabetes, the DCCT trial provides concrete evidence for the cardiovascular benefits of an intensive insulin program.⁹³ Even after the publication of ACCORD, ADVANCE and VADT (Veterans Affairs Diabetes Trial), the question of whether glucose management reduces macrovascular events in type 2 diabetes remains unclear (Tables 28–7 and 28–8).

For type 2 diabetes mellitus, the UKPDS program played a major role in establishing that optimal glycemic targets decreased long-term microvascular complications. Currently, the ADA guidelines target glycemic control at glycated hemoglobin level less than 7% without causing significant hypoglycemia.

With the publication of ACCORD, ADVANCE, VADT, STENO-2 Mortality, and UKPDS-10 years, there is a greater wealth of evidence, but the question of glycemic control and macrovascular outcomes remains confusing.^94,95,96,97

The ACCORD trial is a pivotal randomized study sponsored by the National Heart Lung and Blood Institute (NHLBI), including over 10,000 individuals with type 2 diabetes mellitus and established cardiovascular disease or additional cardiovascular risk factors. In one part of the trial, the comparison between tight glycemic control to keep hemoglobin glycated hemoglobin levels below 6.0% versus conventional control (glycated hemoglobin between 7% and 7.9%) and the reduction of combined cardiovascular events (death by cardiovascular causes, nonfatal MI, and cerebrovascular accident) was investigated. ACCORD was stopped prematurely because of a 22% excess in total mortality in the treatment group at 3.5 years. A higher incidence of hypoglycemia, weight gain, and fluid retention was also observed in the intensive treatment group when compared to the conventional treatment group. The investigators suggest that the higher mortality in the aggressively treated group might be associated with the intensity and time to glycated hemoglobin levels decrease and/or changes in the treatment regimen with changes in drug types and dosages, eventual interaction of the different classes of drugs used in high doses and their side effects, as well as the combination of all of these factors with the clinical characteristics of each patient.

The mechanism for this mortality excess remains unclear, but there was a trend for a 10% reduction in major cardiovascular events overall with lower incidence of nonfatal MI (respectively, 3.6% vs 4.6%, CI 0.62-0.92; P = .004) with aggressive treatment.

Evidence from updated meta-analyses indicates that, for cardiovascular death, there is no difference between the intensive (< 7.0% glycated hemoglobin) and usual care strategies.⁹⁸ In long-term studies that combine UKPDS, ACCORD, and VADT, MI was reduced by approximately 15% over the long term. Apart from this modest reduction in MI rates, there was no significant benefit of a more intensive lowering of glycated hemoglobin with regard to cardiovascular macrovascular end points. One meta-analysis of 27,000 participants with 2370 major cardiovascular events, including hospitalization or death from heart failure, showed no important differences between the more intensive arm compared to the less intensive arm.⁹⁹

Dyslipidemia

Lipid disorders constitute one of the cornerstones in the cardiovascular management of diabetic patients. Many factors influence the lipid profile in these patients, including glycemic control, whether the diabetes is type 1 or type 2, and the presence of diabetic nephropathy.

In type 1 diabetes mellitus, the major determinant of the lipid profile is the level of glycemic control. LDL is moderately increased, triglycerides are markedly increased, and HDL is decreased when the level of glycemic control is impaired. For patients with type 2 diabetes, lipid abnormalities are related not only to hyperglycemia but also to the interplay of the insulin-resistant state. Patients with type 2 diabetes may have normal LDL levels but elevated levels of the very-low-density lipoprotein (VLDL) triglycerides moiety and reduced HDL levels. The expected elevation in VLDL triglyceride is usually no more than 100%.

Low-Density Lipoprotein Cholesterol

Although LDL levels in patients with controlled type 1 or type 2 diabetes may be normal, the atherogenic properties of LDL are increased.¹⁰⁰ There is glycosylation of both apoprotein B and the phospholipid component of LDL, which changes LDL clearance and susceptibility to oxidative modifications. Glycosylation of apoprotein B occurs mainly in the LDL receptor-binding area and is directly related to glucose levels. As a result, there is impairment in the LDL receptor-mediated uptake, and therefore clearance of LDL. Glycosylation also makes LDL more susceptible to oxidative modification. The product generated by the combined glycosylation and oxidation of LDL is more atherogenic than is either glycosylated or oxidized LDL alone. Such LDL molecules are taken up more easily by the aortic intimal cells and macrophages, resulting in the formation of foam cells.

Type 2 diabetic patients with insulin resistance have LDL particles that are small and rich with triglycerides but have little cholesterol in them (small, dense LDL). These LDL particles increase the risk of CHD independent of the total LDL level, probably because of their increased susceptibility to oxidative modification. Therefore, even though LDL levels may be normal in these patients, high levels of small, dense LDL may contribute to the increased risk of CHD in such patients.

High-Density Lipoprotein Cholesterol

A low HDL level is a strong risk factor for the development of CHD in the diabetic patient. There is decreased production and increased catabolism of HDL in diabetes. The decreased HDL production is a result of decreased lipoprotein lipase activity. The failure of lipoprotein lipase to efficiently catabolize VLDL results in reduced availability of surface components for HDL production. By contrast, increased catabolism of HDL results from the hypertriglyceridemia of diabetes, producing triglyceride-rich HDL₂ that is prone to catabolism by liver enzymes.

Dyslipidemia Management

It is recommended that diabetic patients have a lipid profile performed at or around the age of 40 years and every 1 to 2 years thereafter. The frontline primary management strategy remains lifestyle modification with reduction of saturated fats and cholesterol. This specifically involves increase in viscous fiber and plant sterols combined with weight loss.

Low-Density Lipoprotein Lowering

Statin Therapy

Hydroxymethylglutaryl coenzyme A (HMG-CoA) reductase inhibitors—statins—are the frontline therapy in lowering LDL cholesterol levels in type 2 diabetes patients without having an adverse effect on glycemic control.

The management of diabetic patients with lipid abnormalities is a unique challenge to the cardiologist. Important evidence from large randomized trials of lipid-lowering therapies is based on subgroup analyses in which diabetic patients represented less than 10% of all the patients enrolled; however, more recent studies have been done exclusively in diabetic patients. The 4S study enrolled 202 diabetic patients with a prior history of CHD.¹⁰¹ Although this number was too small, the comparison of simvastatin with a placebo showed almost a 50% reduction in coronary events in favor of simvastatin (45% vs 23%; P = not significant). Similar trends were observed in the Cholesterol and Recurrent Events (CARE) Trial, which compared pravastatin with a placebo in secondary prevention. In the CARE trial, the baseline mean LDL concentration in diabetic patients was 136 mg/dL. LDL was reduced 27% in the group receiving pravastatin, which translated into a 25% reduction in coronary events over 5 years compared with that of the control group.¹⁰² The Heart Protection Study (HPS), with a subgroup of 5963 diabetic patients, showed a 28% reduction in total CHD (nonfatal MI and CHD death), nonfatal and fatal strokes, coronary and noncoronary revascularizations, and major vascular events (total CHD, total stroke, or revascularizations) with simvastatin therapy.¹⁰³

In the trials of statin therapy with hyperlipidemia, the relative benefit appears similar between diabetic patients and nondiabetic patients. The CARDS (The Collaborative Atorvastatin Diabetes Study) trial showed that among 2838 diabetic subjects with at least one heart disease risk factor, but without elevated cholesterol levels, randomized to atorvastatin versus placebo and followed up for 3.9 years, statin therapy was associated with a 37% reduction in the primary composite end point of CHD death, fatal MI, hospitalized unstable angina, resuscitated cardiac arrest, coronary revascularization, and stroke.¹⁰⁴

The Deutsche Diabetes Dialysis (4D) Study, in contrast, studied 1255 diabetic patients on maintenance hemodialysis, but did not show a significant reduction in CHD death, fatal MI, or stroke with atorvastatin compared with placebo.¹⁰⁵

Once patients are on statin therapy, LDL determinations are recommended for improving adherence to therapy but not for monitoring the effect of therapy.

• Secondary prevention: For patients with diabetes and known cardiovascular disease, high-intensity statin therapy is recommended according to the American College of Cardiology/American Heart Association (ACC/AHA) guidelines of 2013.¹⁰⁶

• Primary prevention: For patients with diabetes under the age of 40 years with one additional cardiovascular risk factor or those age 40 to 75 years without any cardiovascular risk factors, moderate to high intensity statins are recommended.

In a pooled analysis from the TNT (Treating to New Targets) and IDEAL (Incremental Decrease in Clinical Endpoints Through Aggressive Lipid Lowering) trials, patients with prediabetes treated with high-intensity compared to low-intensity statins were more likely to develop new onset diabetes over 5-years (HR 1.20, 95% CI; 1.04-1.37).¹⁰⁷

Ezetimibe

The recent publication of the Improved Reduction of Outcomes: Vytorin Efficacy International Trial (IMPROVE-IT) trial has brought ezetimibe back into focus as a potential add-on therapy to statins for acute coronary syndrome patients.¹⁰⁸ Of the over 18,000 patients enrolled, 4933 were diabetic and they enjoyed even greater benefit in cardiovascular reduction than nondiabetic patients (ezetimibe in diabetics; HR 0.86; 95% CI, 0.78-0.94), whereas without diabetes, ezetimibe reduced MACE by only ~ 2%; HR 0.98; 95% CI, 0.92-1.04; P-interaction = .02). The use of ezetimibe will likely increase, at least in the acute coronary syndrome population with diabetes.

PCSK-9 Inhibitors

Proprotein convertase subtilisin/kexin type 9 (PCSK-9) antibodies hold the promise of significant isolated LDL-lowering of around 60% and have been approved in the United States (alirocumab, evolocumab) for patients who are unable to reach target LDL levels on statins alone, either for lack of optimal LDL lowering or because of statin intolerance at higher doses.^109,110 They work by binding to LDL receptors, leading to their degradation and reduction in serum LDL levels. At present there are no specific differences in the efficacy and safety of these compounds between diabetic and nondiabetic patients.

High-Density Lipoprotein and Triglycerides

The addition of niacin and fibrates to statin has not been shown to provide any additional cardiovascular benefit in the following trials.

The use of niacin to raise HDL, particularly in diabetic lipid disorders has been attractive. However, in the setting of combination therapy with statins, niacin has failed to improve cardiac end points in two major trials: the Atherothrombosis Intervention in Metabolic Syndrome with Low HDL/High Triglycerides and Impact on Global Health Outcomes (AIM-HIGH) and the Heart Protection Study 2-Treatment of HDL to Reduce the Incidence of Vascular Events (HPS-2 THRIVE).^111,112 There remains some controversy about these negative results: namely, the rather reduced follow-up in AIM-HIGH and the use of laropiprant in combination with niacin in HPS-2.

The hypertriglyceridemia of diabetes can be treated effectively with fibric acid derivatives without an adverse effect on glucose metabolism. These drugs cause a drop of 5% to 15% in LDL levels in patients with normal triglyceride levels, but in patients with hypertriglyceridemia, LDL levels go up. This elevation probably is caused by the catabolism of the atherogenic LDL particle, resulting in less atherogenic LDL. Fibrate therapy also made a lot of sense prior to the reports from pivotal randomized trials. At present, the use of fibrates for isolated hypertriglyceridemia in individual cases remains an option.

The Fenofibrate Intervention and Event Lowering in Diabetes (FIELD) trial studied 9795 patients with diabetes at risk for CHD, and showed that fenofibrate was not associated with a difference in the primary composite end point of CHD death or nonfatal MI compared with placebo at 5 years of follow-up.¹¹³ The treatment effect differences, however, may have been attenuated as a consequence of the more frequent use of statins as lipid-lowering therapy in the placebo arm.

The ACCORD lipid trial did not demonstrate any reduction in cardiovascular events when fenofibrate was combined with simvastatin.¹¹⁴ Fenofibrate in combination with simvastatin did reduce postprandial triglyceride levels, thus raising the question of targeted use in diabetic dyslipidemia.^115,116

Hypertension

The presence of hypertension in diabetic patients significantly increases their risk of micro- and macrovascular complications. It is estimated that 11 million Americans have both diabetes and hypertension. This “deadly duo” definitely march together.¹¹⁷ Furthermore, hypertension in diabetic patients has been linked with numerous other vascular complications, such as nephropathy, retinopathy, the development of cerebrovascular disease, and significant decline in cognitive function in middle-aged diabetic hypertensive patients.

Hypertension Management

Current guidelines suggest the following goals for the management of blood pressure in patients with diabetes: systolic blood pressure of less than 140 mm Hg and diastolic blood pressure less than 90 mm Hg, which represents liberalization compared to the Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC-7) targets.¹¹⁸ This was on the heels of the ACCORD Blood Pressure trial and the (The International Verapamil-Trandolapril Study) subgroup analysis suggesting that a target blood pressure of 140 mm Hg was adequate when compared to more aggressive targets.^119,120

Lifestyle modification is frontline therapy, with weight loss and a Dietary Approaches to Stop Hypertension (DASH) diet remaining the cornerstone of management. Most diabetic patients will require multiple drug therapies with an ACE-inhibitor or ARB included across the board.¹ An important practice-related strategy is to give at least one of the antihypertensives at night. This has conferred additional cardiovascular benefit.¹²¹

The UKPDS demonstrated no advantage of captopril over atenolol in reducing macrovascular complications.¹²² Clearly, this illustrates the significant role that lowering of blood pressure plays in reducing adverse events independent of the agent used. The role of further blood pressure reduction, even when high-risk patients such as diabetic patients are in the normal range, needs to be delineated further. The Hypertension Optimal Treatment (HOT) study showed that the risk of major cardiovascular events in diabetic patients was halved if they had a target diastolic pressure of 80 mm Hg of less compared with those with a diastolic pressure of 90 mm Hg or less (P for trend = .005).¹²³ There was a lower but still significant decrease in the risk of silent MI and approximately a 30% risk reduction in the rate of stroke in the 80 mm Hg or less group compared with the 90 mm Hg or less group.

The HOPE trial evaluated over 9000 high-risk patients with evidence of vascular disease or diabetes in a randomized trial comparing ramipril with placebo over a 5-year period.¹²⁴ A total of 3578 of these patients had diabetes. This study demonstrated a 22% reduction in primary cardiovascular end points of death, MI, and stroke in favor of ramipril. The beneficial effect of ramipril was observed over all predefined subgroups. Interestingly, there was a 30% reduction in the diagnosis of new diabetic patients in the ramipril-treated arm.

In recent years, therapy with dual renin–angiotensin–aldosterone system (RAAS) inhibition has been suggested as a routine practice, particularly in diabetic patients. On the other hand, data from Telmisartan Alone and in Combination with Ramipril Global Endpoint Trial (ON-TARGET), which evaluated a combination of ramipril and telmisartan, demonstrated a worsening of major renal outcomes despite a reduction in proteinuria compared with monotherapy.¹²⁵

Antiplatelet Therapy

For men over the age of 50 years and women over the age of 60 years with at least one additional major cardiovascular risk factor, it is recommended that aspirin therapy at 81 mg daily be instituted as a primary prevention statin strategy for type 1 and type 2 diabetic patients. However, for men and women who are under the age of 50 years and 60 years, respectively, with no additional major risk factors, aspirin therapy is not recommended.¹

Bariatric Surgery

In patients with severe/morbid obesity, bariatric surgical options, such as gastric bypass and gastroplasty, may be appropriate and allow significant improvement in glycemic control with reduction or discontinuation of medications. Commonly used bariatric procedures include the Roux-en-Y gastric bypass with a less invasive gastric sleeve or vertical sleeve gastrectomy (Fig. 28–11).

Emerging evidence from large scale randomized trials from both Mingrone and Schauer indicate that for patients with a BMI of more than 35 kg/m² with type 2 diabetes, bariatric surgery may be a viable option.^126,127 In the Schauer study from the Cleveland Clinic, the glycated hemoglobin target was achieved in only 5% of those on intensive medical therapy compared to 38% for the gastric bypass group and 24% for the sleeve gastrectomy group out to 3 years. Both surgical groups were statistically superior to intensive medical therapy alone with respect to important biomarkers (Fig. 28–12).

Bariatric procedures are not only reducing glycated hemoglobin levels but also the number of antidiabetic drugs required. Approximately 72% of patients treated with bariatric surgery have near or complete stabilization of glycemic control at 2 years in the SOS (Swedish Obese Subjects) study.¹²⁸ Interestingly, the highest rates of remission after bariatric surgery tend to be in patients with higher serum insulin levels with a shorter duration of diabetes.

It is important to fully evaluate the patient for existing or risk for cardiovascular disease and improve glycemic control preoperatively so as to decrease the risk of complications. It is important to counsel patients on the risks of surgery, including mortality, depression, hypoglycemia, nutritional deficiencies, osteoporosis, and weight regain over the long-term. Very little data are currently available on the long-term consequences of surgery for weight loss in people with diabetes. The potential benefits should be weighed against short- and long-term risks.

Newer Therapies

Islet Cell Transplants

Whole-pancreas transplants have successfully restored insulin secretion in people with advanced diabetes but are usually limited to those who are also undergoing kidney transplantation. In 2000, Shapiro and colleagues developed the Edmonton Protocol for islet transplantation, which used a larger quantity of islet cells with drugs that were less toxic to the immune system.¹²⁹ This method infuses islet cells through a small tube into the portal vein of the liver. Patients whose islet cells fail to continue secreting insulin can be retransplanted. Islet cell transplants are still experimental and are available to people who are willing to participate in a study protocol. Also, only a small percentage of islet cell transplant recipients achieve normal blood glucose levels. It is unclear whether the islet transplants will stop or reverse secondary complication related to diabetes. It is also unclear whether islet cell transplantation will ultimately extend a patient’s long-term survival. As of 2008, data from the Collaborative Islet Transplant Registry (CITR) indicates that 71% of adults with type 1 diabetes receiving islet cell transplants were insulin dependent at 1 year, 52% at 2 years, and 23% at 3 years.¹³⁰

Pancreas Transplantation

Ten percent of pancreas transplantations are performed in nonuremic patients with very labile and problematic diabetes.¹³¹ The negative aspect of whole-pancreas transplantation is the need for immunosuppression, which increases the risk of viral and fungal infections and some types of malignancy. Recipients of a pancreas transplant alone have an average 1-year pancreas graft survival rate as high as 78% to 83%.

Successful pancreas transplantation will reverse the thickening of glomeruli and tubular basement membranes, as well as the increase in mesangial volume. Motor sensory and autonomic neuropathy are reversed within 12 to 24 months after transplantation.

Artificial Pancreas

The closed-looped artificial pancreas is being aggressively developed. Russell, Damiano, and colleagues in Boston have shown that compared to an insulin pump, a wearable, automated bionic pancreas delivering both insulin and glucagon was able to improve glycemic targets with less frequent episodes of hypoglycemia over a 5-day period in 20 adults and 32 adolescents with type 1 diabetes mellitus.¹³²

Another collaborative ongoing series of studies led by Kovatchev and collaborators from Charlottesville, Virginia and other centers, uses a smartphone app that receives data from a CGM and gives commands via Bluetooth to an insulin pump. These two devices are being actively studied both in inpatient and ambulatory settings.¹³³

Gene Therapy

Type 1 diabetes mellitus is caused by T-cell–mediated destruction of pancreatic insulin-producing beta cells. Tian and colleagues found a novel way to restore central tolerance in nonobese diabetic mice using hematopoietic stem cells retrovirally transduced to express a protective form of the major histocompatibility complex class II β chain.¹³⁴ As a result, autoreactive T cells will be killed in the thymus and never get to the pancreatic beta cells. Central tolerance refers to mechanisms of tolerance acting in the thymus or bone marrow, in contrast to peripheral tolerance, which occurs in immune cells after they have left the primary lymphoid organs. Preclinical studies must be completed before stem cells can be successfully given to humans with type 1 diabetes mellitus. Some drugs may be synthesized so that they exert their effect only within the areas of inflammation. One example is an engineered transforming growth factor-1β that can become activated locally within areas of beta-cell inflammation.

Mechanical Closed-Loop Sensors

Mechanical closed-loop systems are currently under development.¹³⁵ They consist of a continuous glucose sensor and an insulin pump that can either infuse insulin subcutaneously or directly into the portal circulation. Such a device would not require immune suppression but would be potentially subject to mechanical breakdowns. In theory, these devices could provide both “basal” and synchronized “bolus” insulin requirements. One such device tested in 10 type 1 diabetic patients was implanted without complications. However, the system’s success was limited by pump slowdowns due to insulin precipitation and limited sensor longevity (average of 9 months).¹³⁶

Stem Cells

Stem cells are a potential source of beta cells, which may restore the deficient beta-cell mass found in diabetes,¹³⁷ and are the “holy grail” for the treatment of type 1 diabetes. This concept would assume the source of beta-cell destruction could be abated. Ianus and colleagues found an extrapancreatic source of pancreatic beta cells in bone marrow–derived cells in a mouse model. Pancreatic endocrine and exocrine cells both originate from epithelial cells from early gut endoderm. A second source of stem cells is in the pancreas from progenitor epithelial cells in the pancreatic duct. Much more work is required to translate the potential that stem cells have to produce insulin on demand clinically.

Stem cell therapy is in the research pipeline and the reader is referred to several articles for more data.^138,139,140

A central theme of current research is that inhibition of tumor growth factor-β signaling an adult murine beta cells can promote beta-cell replication.¹⁴¹ Stem cell differentiation and transdifferentiation from alpha or delta cells may be possible. Also, autoimmunity needs to be harnessed. We must stay tuned for further developments in this exciting area.

Coronary Revascularization

The management of diabetic patients with CHD entails both pharmacologic and revascularization strategies. Over the past several years, there have been many advances in the medical management of the diabetic patient with CHD. Aspirin, β-blockers, statins, and ACE inhibitors are routinely administered. These agents may provide clinical benefit not only by treating ischemia, but also by stabilizing atherosclerotic plaque and inhibiting endovascular thrombosis, thereby preventing acute coronary events.

Coronary revascularization procedures have become a mainstay of therapy for CHD patients, providing both symptomatic relief and mortality reduction in certain anatomic subsets. Several studies have attempted to rationalize the use of different revascularization techniques by comparing them to medical therapy and to each other in various clinical settings.

The evaluation of coronary revascularization in diabetic patients was enriched by a number of important trials: CARDIA (The Coronary Artery Revascularisation in Diabetes), VA-CARDS,^142,143 SYNTAX (Synergy between PCI with Taxus and Cardiac Surgery), and FREEDOM (Future Revascularization Evaluation in Patients with Diabetes Mellitus: Optimal Management of Multivessel Disease). The SYNTAX program consisted of a randomized trial of patients with three-vessel or left main disease comparing PCI with paclitaxel-eluting stents versus coronary artery bypass graft (CABG) as well as two parallel registries of patients ineligible for PCI or CABG respectively. In the SYNTAX trial about 25%, or 452 of the 1800 patients, had diabetes.¹⁴⁴ At 12 months of follow-up, the only difference between PCI and CABG for diabetic patients was the excess of repeat revascularization in the PCI arm (20.3% vs 6.4%; P < .001). As expected, when compared to nondiabetic patients, the 12-month rate of death, MI, or stroke was higher for the diabetic cohort (10.2% vs 6.8%). Interestingly, when moving across the spectrum of non-diabetes to metabolic syndrome to insulin-treated diabetes, the 12-month major adverse cardiovascular and cerebrovascular events (MACCE) rate for CABG was consistent at between 12% and 14%, whereas, the MACCE rates for PCI progressed from 15% in nondiabetes to 30% in insulin-treated diabetes. The SYNTAX investigators intend to follow the patients for up to 5 years.

Another important aspect of the SYNTAX program was the derivation and evaluation of the SYNTAX scoring system for grading the complexity of CAD. Simply, the SYNTAX score is the summation of the points assigned for each individual lesion in the coronary tree (divided into 16 segments). The percent stenosis is not included, but rather the scoring is assigned according to occlusive versus degrees of nonocclusive disease, and the complexity of a given lesion (eg, bifurcation lesion) is factored into the score. Based on angiographic criteria, the investigators demonstrated that patients with higher scores and, therefore, more complex disease were more likely to benefit from CABG surgery over PCI. This is based on the observation that not all three-vessel and left main disease are equivalent.

The CARDIA trial evaluated diabetic patients with multivessel disease exclusively to either drug-eluting stent (DES)-PCI or CABG. This 500-patient trial was underpowered to detect important clinical differences in death, MI, or stroke at 12 months. In a noninferiority analysis, the primary event rate was greater in the PCI arm (11.6 vs 10.2%), but this failed to meet statistical significance. PCI was not proven to noninferior to CABG because of wide 95% confidence intervals. As in SYNTAX, there was an excess in repeat revascularization in the PCI arm. CARDIA set the stage for the FREEDOM trial, which will have 1900 subjects with a mean follow-up of 3.8 years. The FREEDOM investigators, in 2012,¹⁴⁵ reported that at 5 years the primary outcome (composite of all-cause mortality, nonfatal MI, or nonfatal stroke) was more likely to occur in the PCI group than in the CABG group (26.6% vs 18.7%, respectively; P = .005). The percent-year rates of nonfatal MI and all-cause mortality were greater in the PCI arm (13.9% vs 6.0%; P < .001 for nonfatal MI and 16.3% vs 10.9%; P = .049 for all-cause mortality), and the rate of nonfatal stroke was significantly lower in the PCI-DES arm (2.4% vs 5.2%; P = .03). Given the findings from COURAGE (Clinical Outcomes Utilizing Revascularization and Aggressive Drug Evaluation) and from BARI 2D (The Bypass Angioplasty Revascularization Investigation 2 Diabetes), FREEDOM had strict medical management targets for both strategies. FREEDOM led to a class 1 indication for CABG in diabetic patients with multivessel CAD when using the heart team approach.¹⁴⁶

See Chap. 44, which reviews the trials of PCI versus CABG in detail.

Newer Generation Drug-Eluting Stents

By the time that FREEDOM was reported, we witnessed the advent of second- and later generation DES platforms. New evidence suggests that later generation stents have closed the gap between CABG and PCI in diabetic patients.¹⁴⁷ The TUXEDO (The Taxus Element versus Xience Prime in a Diabetic Population) trial and meta-analysis that included patients from 68 randomized controlled trials showed, for example, using cobalt-chromium everolimus-eluting stents may be promising.¹⁴⁸ From BARI to ARTS to FREEDOM, the point estimate for mortality remains unchanged, so it would be a mistake to continue to test the CABG versus PCI hypothesis without first demonstrating that systemic therapy with optimal medical management can be improved to the point where nontarget lesions are no longer an issue.

Comprehensive Meta-Analysis of PCI versus CABG

One major meta-analysis by Hlatky and colleagues from the early percutaneous transluminal coronary angioplasty, or PTCA, and bare metal stent (BMS) era reported on their findings. The diabetes subgroup appeared to fare better with CABG when 5-year findings were pooled.¹⁴⁹ Verma and coworkers reported on the 5-year survival benefit of CABG over PCI in diabetic patients with multivessel CAD showing little difference between BMS-PCI and DES-PCI. The increased risk of stroke after CABG was significant and consistent with the FREEDOM findings.¹⁵⁰

Screening Asymptomatic Diabetic Patients—DIAD

The Detection of Ischemia in Asymptomatic Diabetics (DIAD) trial publication answered an important question in the diabetes and CAD field: namely, what is the role of routine screening for myocardial ischemia in asymptomatic diabetic patients?¹⁵¹ For many years, the American Heart Association had advocated for optimal management of coronary risk factors as frontline therapy. Colleagues from the ADA felt strongly for routine stress testing in these patients. Over 1000 patients asymptomatic patients were randomized to undergo myocardial perfusion imaging or no screening. After 5 years, there was no difference in the rates of nonfatal MI or cardiac death between the two arms (relative risk reduction 12%; 95% CI, 78%-56%). The DIAD trial demonstrated a very low rate of significant ischemia in the screening arm (about 6% with moderate to large perfusion defects) and a very high adherence to a robust medical risk factor modification program. In fact, the excellent medical therapy in DIAD is responsible for a very low 5-year primary event rate. The National Institutes of Health-funded ISCHEMIA (International Study of Comparative Health Effectiveness With Medical and Invasive Approaches) trial aims to determine what the best therapy for patients with moderate to large perfusion defects is if they a treated aggressively, in an adequately powered trial. If there is no benefit to revascularization in these patients, then there is no rationale to screen for ischemia. Further analysis from BARI 2D may help address this issue.

Coronary Revascularization Compared to Optimal Medical Therapy in Less Aggressive CAD

About one-third of the patients randomized in the COURAGE trial were diabetic.¹⁵² The trial demonstrated no difference in the 5-year MACE rate. This required further confirmation. BARI 2D was a multicenter NHLBI-sponsored study to address two important question.¹⁵³ First, it addressed the question of whether diabetic patients with relatively mild or no symptoms could be treated with deferred (optimal medical therapy [OMT] alone) compared with prompt revascularization on top of OMT. The important feature was that patients were randomized postcatheterization with anatomy defined and required at least one lesion to be amenable for revascularization. It is critical to emphasize that the decision to undertake PCI or CABG was left to the discretion of the investigators with the caveat that patients with multivessel and more extensive disease were more likely to undergo CABG. Second, BARI 2D addressed the question of whether treatment with an insulin-sensitizing strategy was superior to a strategy of insulin provision with a target glycated hemoglobin of less than 7.0% in both arms. Patients with left main disease or creatinine greater than 2.0 mg/dL were excluded.

In the end, 2368 patients were enrolled at 60 international sites with 1605 in the PCI versus OMT strata and 763 in the CABG versus OMT strata (see Fig. 28–12). As expected, 93% of patients with single-vessel disease were enrolled in the PCI stratum and, conversely, 55% of triple-vessel disease patients were enrolled in the CABG stratum. The primary outcome measures of the trial demonstrated no difference in 5-year mortality for either of the comparison of OMT versus prompt PCI (12.2% vs 11.7%; P = .97) or the comparison of an insulin-sensitizing versus an insulin-providing strategy (11.8% vs 12.1%; P = .89). Similarly, the rates of MACE between the two principle comparisons did not differ.

BARI 2D was an important contribution because two key conclusions can be drawn. One, patients with diabetes mellitus who are asymptomatic or have mild symptoms can be first treated with OMT alone. By 5 years, 40% of these patients will require a deferred revascularization procedure. Two, when OMT is applied, the target of glycated hemoglobin of 7% can be achieved by an individualized approach. Patient adherence, polypharmacy, weight gain, and edema are all factors that will factor into the antidiabetic strategy employed. The TZDs were safe in BARI 2D, with a high prescription of rosiglitazone observed.

Increasing evidence suggests that an atherogenic prediabetic state exists prior to the development of diabetes. In the San Antonio Heart Study, patients converting to overt diabetes had significantly higher blood pressure, BMI, waist circumference, triglyceride levels, and lower HDL cholesterol levels at baseline compared to nonconvertors.¹⁵⁴

Investigators from Finland conducted a randomized trial of 522 middle-aged, overweight subjects with impaired glucose tolerance (350 women, mean age: 55 years).¹⁵⁵ Patients were randomized to a control group or an intervention group that consisted of individualized counseling aimed at reducing weight, total intake of fat, and intake of saturated fat, and increasing intake of fiber and physical activity. Over 4 years of follow-up, the cumulative incidence of diabetes was 11% in the intervention group and 23% in the control group (58% reduction; P < .001). The authors concluded that type 2 diabetes could be prevented with an intensive lifestyle modification program.

The Diabetes Prevention Program evaluated more than 3000 nondiabetic subjects with impaired fasting, and postglucose loading plasma glucose levels and showed that when compared to placebo, a lifestyle intervention reduced the incidence of developing diabetes by 58% (95% CI interval, 48%-66% and NNT of 7).¹⁵⁶ Metformin reduced the incidence by 31% (95% CI interval, 17%-43%; NNT of 14). When compared head to head, the intervention group fared better than the metformin group.

Trials with antidiabetic agents in preventing type 2 diabetes have had limitations: in the Diabetes Reduction Assessment with Ramipril and Rosiglitazone Medication (DREAM) trial 5269 patients with either impaired glucose tolerance or impaired fasting glucose were randomized in a factorial design to ramipril and/or rosiglitazone to determine whether such therapy prevents the onset of new type 2 diabetes. Rosiglitazone therapy led to a highly significant reduction in incident diabetes (HR 0.04, 95% CI 0.35-0.46; P < .0001); however, the use of rosiglitazone has fallen off because of safety concerns.^157,158 The Nateglinide and Valsartan in Impaired Glucose Tolerance Outcomes Research (NAVIGATOR) trial, the largest diabetes prevention clinical trial to date, showed a modest reduction of 14% in the development of type 2 diabetes with valsartan, but not nateglinide, in 9150 subjects who had impaired glucose tolerance and were at high cardiovascular risk.^159,160

The significant increase in major microvascular and macrovascular complications makes it important to begin screening for diabetes at an age younger than 45 years.¹⁶¹ It has become necessary to implement aggressive screening strategies to be able to identify populations at the highest risk of developing diabetes.¹⁶²

Current measures of cardiovascular surveillance for CAD in asymptomatic diabetic patients focus on routine stress testing in accordance with the ACC/AHA guidelines. Exercise testing in diabetic patients is more likely to be accurate when combined with echocardiography or radionuclide imaging.¹⁶³ Diabetic patients are less likely to have an appropriate blood pressure and heart rate response to exercise and less likely to experience any pain corresponding to ST-segment changes caused in part by autonomic dysfunction. The AHA recommends that the finding of subclinical CAD should prompt clinicians to initiate more aggressive preventative measures. The DIAD study has shown that the prevalence of silent ischemia in the diabetic population is not insignificant, but the annual cardiac event rate is less than 1% overall at 4.8 years of follow-up and that routine screening for inducible coronary ischemia did not reduce cardiovascular events.¹⁶⁴

Novel Strategies for Early Detection and Risk Prediction

There are a number of strategies aimed at identifying subclinical atherosclerosis in diabetic patients. Two strategies that have the best evidence include coronary artery calcium (CAC) and ankle brachial index (ABI) determination. CAC imaging has been shown to provide incremental predictive value over traditional cardiovascular risk factors.¹⁶⁵ The ABI is a cost-effective technique for screening that is underutilized, and three-dimensional carotid ultrasound and not carotid media thickening is showing promise in large cohort studies.¹⁶⁶ Other strategies, such as measuring for cardiac autonomic dysfunction through assessment of heart rate variability and QT index, may provide incremental prognostic value in predicting mortality.¹⁶⁷

On the clinical front, there are still many challenges in the prevention and management of diabetic cardiovascular complications. The focus on a multidisciplinary approach to diabetic cardiovascular disease is critical because we are dealing with a chronic, systemic illness. Glycemic control appears to be the mainstay of long-term diabetes management. Thus, development of better therapies and devices (eg, gene therapy, stem cell therapy, artificial pancreas, closed-loop pumps, islet and pancreatic transplants) for achieving and maintaining glycated hemoglobin at not only less than 7%—but in the normal range of less than 6%—will be a primary goal in the next decade. The findings of the first antidiabetic trial to demonstrate a benefit both in cardiovascular mortality and heart failure hospitalization, EMPA-REG, were recently reported and this trial is a potential transformational development. The results of cardiovascular outcomes trials with SGLT-2 agents are anxiously awaited. The superiority of CABG over PCI with DES in FREEDOM led to a class 1 recommendation for the treatment of multivessel CAD. Strategies to reduce the increased risk of periprocedural stroke in CABG remains an important challenge for the next decade as the diabetic population with CAD ages and a focus on quality of life will prevail. Significant improvement in OMT with aggressive LDL-lowering agents will potentially alter the need for coronary revascularization down the road. Finally, the role of cell-based therapies in the management of diabetic atherosclerotic vascular disease needs to be addressed within the context of all other advances.