A 58-year-old man with a history of hypertension, diabetes, and hyperlipidemia was admitted to the hospital because of a non-ST-segment elevation myocardial infarction. He underwent successful percutaneous coronary revascularization, including placement of a drug-eluting stent in his left circumflex artery. He was discharged on a regimen of aspirin, clopidogrel, metoprolol, lisinopril, and simvastatin.

 One week following hospital discharge, the patient returned to work as a construction worker and was able to resume strenuous work without symptomatic limitations; however, 8 months later, he developed sudden chest pain while at work. The pain was similar in nature to what he experienced with his initial myocardial infarction (MI). He was taken to the hospital via ambulance, and a diagnosis of ST-segment elevation MI was made. The patient underwent primary percutaneous coronary intervention for thrombosis of the left circumflex artery stent. Discussions with the patient revealed that in the months following his initial MI, he became depressed because of marital problems and financial stress from having only intermittent work. He also said that it was not clear to him how long he needed to take all of the medications prescribed to him after his first MI. As a result, he cut back on his cardiac medications and recently stopped taking all of them, except aspirin.

 This case report highlights the clinical scenario of recurrent cardiovascular events after stopping cardiac medications, reinforcing the importance of medication adherence. Whether the recurrent cardiac event is directly related to cessation of cardiac medications is unclear, but several studies support that cessation of certain medications (eg, beta blockers and oral antiplatelets such as clopidogrel) is associated with adverse outcomes. This case also shows that the first sign of medication nonadherence may be presentation with a secondary cardiac event; thus, an important clinical implication is the need to incorporate assessment of medication nonadherence into routine clinical practice. This would allow clinicians to potentially detect nonadherence before the occurrence of adverse events, enabling them to work with patients to develop solutions to help them with medication adherence. 

Articles

Cardiology Review Online

He underwent successful percutaneous coronary revascularization, including placement of a drug-eluting stent in his left circumfl ex artery. He was discharged on a regimen of aspirin, clopidogrel, metoprolol, lisinopril, and simvastatin.

The dangers of nonadherence to cardiovascular medications

 Street, PO Box 4088, Frederick, MD 21705.

 Contact: The Cleveland Clinic Center for

Calendar

The fasting lipid profile plays an integral role in preventive cardiology. By obtaining total cholesterol, triglyceride, high-density lipoprotein (HDL) cholesterol, and low-density lipoprotein (LDL) cholesterol levels, we are provided a snapshot of the atherosclerotic risk of our patients and can use these values as targets for therapy directed by national guidelines.

 Yet, the need for a return visit after an overnight fast remains a considerable inconvenience for both patients and providers.

Traditionally, we have requested lipid assessment in the fasting state for 2 basic reasons. First, fasting triglycerides can be used to calculate LDL cholesterol using the Friedewald equation,

 and LDL cholesterol is a well-established target for therapy.

 Second, in contrast to total cholesterol and directly measured LDL cholesterol and HDL cholesterol, triglycerides obtained in the nonfasting state can vary substantially based on the content of a recent meal. Hence, fasting triglycerides are thought to provide a more reliable estimate of cardiovascular risk.

 of their study of nonfasting triglycerides in the Copenhagen City Heart Study.

 They followed 7587 women and 6394 men for a mean of 26 years for cardiovascular end points and mortality. In age-adjusted analyses, nonfasting triglycerides were strongly associated with myocardial infarction (MI), ischemic heart disease, and all-cause mortality; in multivariate analyses, these associations were attenuated but persistent.

This study underscores the potential benefits of measuring nonfasting lipids. However, a major limitation of the study is that the multivariable model does not account for HDL cholesterol, which is strongly and negatively correlated with triglycerides. Previous studies have demonstrated that adjustment for HDL cholesterol substantially reduces the association between triglycerides and cardiovascular disease risk.

 Although the authors suggest that multifactorial adjustment “masks the effects of triglycerides,” the true clinical utility of a novel risk factor is its ability to provide additional, independent risk information beyond established risk factors.

Another recent study also analyzed the effect of nonfasting triglycerides on cardiovascular disease. Bansal and colleagues

 reported the results of more than 26,000 women followed for 11 years for MI, ischemic stroke, coronary revascularization, and cardiovascular death. In multivariable analyses adjusted for traditional risk factors as well as total cholesterol, HDL cholesterol, and markers of insulin resistance, fasting triglycerides were not independently associated with the primary end point; however, women in the top third of nonfasting triglycerides had nearly twice the risk of cardiovascular events as those in the bottom third. In secondary analyses, triglycerides were most associated with cardiovascular risk in women who had eaten 2 to 4 hours prior to their blood draw, a time frame that corresponds to the peak in triglyceride levels after a meal.

The association of nonfasting lipids with cardiovascular disease can be explained by 2 phenomena. First, pathophysiologically, elevated postprandial triglyceride levels signify a large burden of triglyceride-rich atherogenic remnant lipoproteins, particles that have been strongly implicated in atherosclerosis.

 Second, the exaggerated triglyceride response to a meal may be indirectly associated with cardiovascular risk because it is a manifestation of insulin resistance and the metabolic syndrome.

The results of these studies have provocative clinical implications for the practice of assaying lipids after an overnight fast. Direct measurement of LDL cholesterol is now widely available, and in contrast to calculated LDL cholesterol, it does not change substantially after a meal. Furthermore, both apolipoprotein B, a measure of the total number of atherogenic particles, and non-HDL cholesterol, already a secondary target in national guidelines, are stable regardless of prandial state and are established risk factors for atherosclerosis.

 Finally, the observation in these studies that the nonfasting triglyceride level is the stronger independent predictor of cardiovascular risk than the fasting triglyceride level obviates the reason for measuring lipids in the fasting state.

So can we finally begin measuring nonfasting lipids in all our patients? Unfortunately, before the results of these epidemiologic studies can be translated into clinical practice, further studies are needed to identify clinical cutpoints for triglycerides after a standardized meal. Based on these recent studies, however, clinicians might not be so cavalier about disregarding high triglyceride values because they were not reliably measured in the fasting state. And as further evidence accrues regarding the importance of nonfasting lipids, we may someday, at long last, do away with the fasting lipid profile.

The fasting lipid profile plays an integral role in preventive cardiology.

The fate of the fasting lipid profile

	Case report: A patient with cardiogenic shock and impaired fasting glucose

	A 63-year-old man with a history of hypertension was admitted to the coronary care unit with sudden-onset severe chest pain and the clinical signs of severe heart failure. The electrocardiogram showed marked ST-segment elevation in leads I, aVL, and V1 through V6; ST-segment depression in leads II and III; and ventricular fibrillation, suggesting acute anterior myocardial infarction. The patient’s systolic blood pressure was less than 90 mm Hg despite the fluid challenge associated with clinical signs of hypoperfusion, indicating cardiogenic shock. An emergency coronary angiogram showed a tight lesion (99% stenosis) in the left anterior descending coronary artery. Primary angioplasty was performed successfully with stent implantation. Following the procedure, the patient’s hemodynamic characteristics stabilized. Results of laboratory tests showed a peak plasma creatine kinase level of 9,756 UI/L and a mean fasting blood glucose level (taken on days 4 and 5) of 6.6 mmol/L, indicating impaired fasting glucose. n

Case report: A patient with cardiogenic shock and impaired fasting glucose

	Impaired fasting glucose and cardiogenic shock

	Marianne Zeller, PhD1  •  Marie-Claude Brindisi, MD2  •  Gilles Dentan, MD3

	Yves Laurent, MD4  •  Luc Janin-Manificat, MD5  •  Isabelle L’Huillier, MD1

	Jean-Claude Beer, MD1  •  Hamid Makki, MD6  •  Jean-Eric Wolf, MD1

	Bruno Verges, MD, PhD2  •  Yves Cottin, MD, PhD1  •  on behalf of the RICO survey working group

	From the Service de Cardiologie1 and Service d’Endocrinologie,2 CHU Bocage,

	Dijon; Service de Cardiologie, Clinique de Fontaine,3 Fontaine les Dijon; Service de Cardiologie, Centre Hospitalier,4 Semur en Auxois; Service de Cardiologie, Centre Hospitalier,5 Beaune; Service de Cardiologie, Centre Hospitalier, Châtillon sur Seine,6 France

	A new definition for impaired fasting glucose has been established by

	the American Diabetes Association (ADA). Patients are considered to

	have impaired fasting glucose if they have glucose levels from 6.1 mmol/L to 7 mmol/L (110 mg/dL to 126 mg/dL).1 It has been shown that patients with abnormal glucose metabolism are in greater danger of developing macrovascular complications and coronary artery disease (CAD).2 Few

	studies have been conducted on the effects of impaired fasting glucose in patients with acute myocardial infarction (MI), however. The importance of assessing risk in patients with abnormal glucose metabolism was highlighted in a recent study, which showed that patients with CAD who underwent percutaneous coronary intervention had a greater risk of mortality if they had abnormal glucose metabolism.3

	We compared the in-hospital outcome of MI patients with normal fasting glucose versus patients with impaired fasting glucose from the unselected population of the French regional survey for acute MI, observatoiRe des Infarctus de Côte-d’Or (RICO).

	A total of 999 patients from the French regional study hospitalized for acute MI were included in the study. MI diagnosis was defined based on the criteria of the European Society of Cardiology and the American College of Cardiology.4 We did not include patients in the study if their fasting blood glucose levels had not been taken.

	Glycosylated hemoglobin (A1c) levels were taken on the first morning

	following admission to the hospital. Fasting blood glucose levels were taken on days 4 and 5. Using the Cockcroft-Gault formula, the baseline serum creatinine clearance rate was estimated, and the peak plasma level of creatine kinase (CK) was assessed.5

	Patients were considered to have diabetes mellitus if they had a history of diagnosed diabetes mellitus or a mean fasting blood glucose level of 7.0 mmol/L or greater, based on the ADA’s definition of diabetes mellitus. Also using ADA’s definitions, if patients had a mean fasting glucose of 6.1 mmol/L or more (110 mg/dL) but less than 7.0 mmol/L, they were considered to have impaired fasting glucose; normal fasting glucose was defined as a mean fasting glucose less than 6.1 mmol/L.1

	Characteristics and outcome. Of 526 patients (53%) who had abnormal glucose metabolism, 381 (38%) had diabetes mellitus and 145 (15%) had impaired fasting glucose. Patients with diabetes mellitus and impaired fasting glucose had greater peak

	CK plasma levels and greater body mass index than those with normal fasting glucose levels (P < .01; table 1). Compared with the normal fasting glucose group, the impaired fasting glucose group had a marked-

	and mortality (P = .049 and P = .011, respectively).

	Predictors of mortality. The in-hospital mortality was primarily relatedto cardiogenic shock (odds ratio [OR], 4.86; 95% confidence interval [CI], 1.53—15.39; P = .007) and ventricular arrhythmias (OR, 3.62; 95% CI, 1.18–

	11.11; P = .014) in the impaired fasting glucose group, according to multivariate analysis. Table 2 shows that, among patients without diabetes mellitus, impaired fasting glucose was a predictive factor for mortality in univariate analysis (P = .0324), but it was not an independent predictive factor for in-hospital death, after adjusting for anterior location, left ventricular ejection fraction, age, reperfusion, ventricular arrhythmia, sex, and cardiogenic shock. Impaired fasting glucose also had no predictive value for ventricular arrhythmias when adjusted for other variables. In contrast, impaired fasting glucose had a strong predictive value for cardiogenic shock (P = .0217), even after adjustment for confounding factors (P = .027).

	Based on the results of our study, in-hospital mortality is twice as high in patients with impaired fasting glucose compared with those with normal fasting glucose, mainly because of the increased risk of cardiogenic shock.

	Among the unselected MI patients, we found a high preponderance of patients (53%) with abnormal glucose metabolism, with 15% of patients having impaired fasting glucose. In the National Health

	and Nutrition Examination Survey (NHANES) III study, only 7% of healthy adults had impaired fasting glucose.6 In that study, 8% of patients with impaired fasting glucose had CAD, compared with only 6% of patients in the normal fasting glucose group (relative risk [RR], 1.47). In another study, abnormal fasting glucose metabolism was more prevalent than expected in 1,612 patients with CAD undergoing percutaneous coronary intervention.3 These findings are similar to those in our study, which went one step further by including patients with acute MI.

	Previous studies have pointed out the strong predictive value of fast-ing glucose levels in CAD patients for all-cause mortality at long-term follow-up, even after adjustment for potential confounding variables.7 In acute MI patients with admission

	or fasting hyperglycemia (glycemia threshold range, 6.1 to 8 mmol/L), Capes and colleagues found a higher risk of in-hospital mortality (RR, 3.9; 95% CI, 2.9—5.4) for hyperglycemia.8 A higher risk of in-hospital death was associated with impaired fasting glucose in patients with acute MI, which highlights the importance of using the new definitions of abnormal glucose levels for patients with acute MI, as well as the need to determine specific therapy during the acute phase of MI.

	Because impaired fasting glucose was not a persistent prognostic factor for death after adjusting for covariates, especially for cardiogenic shock, it is believed that the increased risk of death associated with impaired fasting glucose was mainly the result of a high occurrence of cardiogenic shock. In patients with MI who have blood glucose levels

	above 10 mmol/L, without adjusting for potential covariates, previous studies have shown a greater risk of severe pump failure.9,10 After adjusting for age, nondiabetic patients with abnormal glucose metabolism have also been found to have a greater risk of congestive heart failure or cardiogenic shock.11

	of cardiogenic shock in patients with impaired fasting glucose are not well understood, although it is theorized that increased glucose levels are involved because high blood glucose levels have been found to be damaging to cardiomyocytes.12 High blood glucose levels also indicate

	a relative insufficiency of insulin, which correlates with greater lipolysis and increased circulating free fatty acids, which may harm cardiac cells.13

	Study limitations. We measured fasting glucose levels in patients with acute ischemic stress and disturbed metabolic state, which may interfere with glucose metabolism. We cannot rule out the fact that increased blood glucose levels may be the result of acute metabolic stress following acute MI. A previous study of nondiabetic patients that measured serial glucose concentrations, however, showed that the fasting blood glucose level after acute MI was stable from day 4 and was an independent predictor for abnormal glucose metabolism at the 3-month follow-up (P < .017). These findings indicate that the increased glucose levels are not just caused by the stress related to an acute MI.14 The A1c level, which is an indicator of preexisting glucose metabolism, in our study was also markedly greater in patients with impaired fasting glucose, thus confirming the presence of abnormal glucose metabolism before the acute MI.

	Our study showed that in-hospital mortality was significantly increased as a result of a higher risk

	in acute MI patients with impaired fasting glucose. This highlights the importance of assessing specific therapy during the acute stage of MI for patients with impaired fasting glucose. n

Impaired fasting glucose and cardiogenic shock

	There is a growing consensus that treatment of the total lipid profile can reduce the burden of coronary artery disease (CAD) more effectively than lowering low-density lipoprotein (LDL) cholesterol alone.1-3 However, no single lipid-altering agent effectively modifies all lipid abnormalities. The HMG-CoA reductase inhibitors (statins) are the most powerful agents for lowering LDL cholesterol, but they have lesser effects on triglycerides and high-density lipoprotein (HDL) cholesterol levels. Niacin is the most effective agent for raising HDL cholesterol, and it further complements statin therapy by effectively lowering triglycerides and lipoprotein (a). Their combination has the advantage of controlling the four major components of the lipid profile influencing the development of atherosclerosis. Moreover, preliminary data suggest that this combination may produce substantial clinical benefits; in the HDL Atherosclerosis Treatment Study (HATS), CAD patients treated with simvastatin plus niacin experienced a 60% to 90% reduction in cardiovascular events compared with placebo.4

	Extended-release (ER) niacin/lovastatin is the first combination agent approved for the treatment of primary hypercholesterolemia and mixed dyslipidemia.5 The present study compared the lipid-altering efficacy of coadministration of niacin ER and lovastatin with niacin ER alone and lovastatin alone using all clinically reasonable doses.6 The dose-response effects were also analyzed by innovative surface graphs.

	The study results have been previously published.6 In brief, this was a 26-week, randomized, double-blind, active-control study performed at multiple clinical centers in the United States. Men and women with type IIa or IIb hyperlipidemia were eligible if they had stable baseline LDL cholesterol levels of 130 mg/dL or higher with CAD or risk equivalents, 160 mg/dL or higher LDL cholesterol with two or more CAD risk factors, or 190 mg/dL or higher LDL cholesterol with fewer than two risk factors and fasting triglyceride levels less than 800 mg/dL during screening.

	Patients were randomly assigned to one of five treatment arms: three arms tested coadministration of niacin ER/lovastatin, one tested niacin ER alone, and one tested lovastatin alone for 20 weeks. In each combination group, the lovastatin dose was fixed at 10, 20, or 40 mg, whereas the niacin ER dose was titrated by 500-mg increments every 4 weeks, up to 2,500 mg. In the monotherapy groups, the niacin ER dose was titrated similarly, whereas the lova-statin dose was doubled every 8 weeks, from 10 mg up to 40 mg. In total, this design included 23 different doses. Patients were counseled to take study medications once daily at bedtime with a low-fat snack. In addition, to reduce flushing, 325 mg of aspirin 30 minutes before dosing was recommended.

	The two primary end points of efficacy were the mean percentage change in LDL cholesterol level for patients taking 1,500/20 mg niacin ER/lovastatin compared with those taking 20 mg of lovastatin alone

	and the mean percentage change in LDL cholesterol for patients taking 2,000/40 mg of niacin ER/lovastatin compared with those taking 40 mg of lovastatin alone. Secondary end points of efficacy included mean percentage change in HDL cholesterol, triglyceride, and lipoprotein (a) levels. Blood was drawn to determine lipid levels every 4 weeks prior to scheduled dose escalation. Treatment-emergent adverse events and serum levels were evaluated

	at each visit. For use in clinical practice, results are given for the FDA-approved niacin ER/lovastatin dosing range of from 500/20 mg to 2,000/40 mg. Three-dimensional surface graphs (figure) show all the results of the study, from 500/10 mg to 2,500/40 mg.

	A total of 164 patients were randomly assigned to one of the five treatment groups. Baseline characteristics were similar among the groups (table 1).6 Medication compliance was 96%, 96%, and 92% with the combination, lovastatin alone, and niacin ER alone, respectively.

	Efficacy. Niacin ER/lovastatin produced favorable, dose-dependent changes in all four components of the lipid profile. For the two pri-mary end points, niacin ER/lova-statin at 1,500/20 mg and 2,000/40 mg lowered mean LDL cholesterol levels more effectively than lova-statin alone at 20 mg and 40 mg, respectively (P < .001 for both comparisons; table 2). Niacin ER/lova-statin at doses of 500/20 mg to 2,000/40 mg produced progressive decreases in LDL cholesterol of 29% to 46% from baseline. These reductions were approximately additive relative to those achieved with the individual components. Each combination of niacin ER (500 mg—2,000 mg) with lovastatin 20 mg also lowered LDL cholesterol more effectively than lovastatin alone at the 40-mg dose. In addition, LDL cholesterol decreased by approximately 4% for each 500-mg increase in the niacin ER dose.

	For secondary end points, niacin ER/lovastatin at doses of 500/20 mg to 2,000/40 mg produced progressive increases in mean HDL cholesterol levels of 7% to 29% and decreases in triglycerides of 9% to 39% from baseline (table 1). The effects of combination treatment were generally additive. In addition, HDL cholesterol increased by approximately 8% for each 500-mg increase in the niacin ER dose. Lipoprotein (a) levels decreased progressively by 5% to 14% with combinations containing niacin ER doses of  1,000 mg or more.

	The overall strength and consistency of the dose-response relationships for each lipid are shown in the surface graphs in the figure.6 They display the incremental efficacy on the four lipid components simultaneously across the 23 doses tested. The figure caption provides a detailed explanation of how to read the graphs.

	Safety. Flushing was the most common side effect seen in patients treated with niacin ER regimens (62%), but it led to withdrawal in only 6% of patients on combination therapy and 10% of patients on niacin ER alone. Excluding flushing, the incidence of adverse events and withdrawals was similar among all five treatment groups. At recommended niacin ER/lovastatin doses less than 2,000/40 mg, one patient had transient increases in liver function test levels greater than three times normal. Three additional patients had increased liver function test levels with combination doses containing 2,500 mg of niacin ER.

	No increases in creatine kinase levels greater than three times normal occurred. No cases of myopathy were reported.

	Optimal control of dyslipidemia is essential for the primary and secondary prevention of CAD. This study shows the complementary and incremental efficacy of niacin ER/

	lovastatin on four lipid parameters independently correlated with CAD risk. In the recommended dosing range of 1,000/20 mg to 2,000/40 mg per day, niacin ER/lovastatin reduced LDL cholesterol levels by 32% to 46%, triglycerides by 24% to 38%, and lipoprotein (a) levels by 5% to 14%, and increased HDL cholesterol levels by 18% to 29%. Niacin ER/

	lovastatin produced generally additive effects on LDL cholesterol, HDL cholesterol, and triglycerides. At the 2,000/40-mg dose, changes in lipids were similar to or greater than those seen in HATS.4 For LDL cholesterol lowering, combination therapy was more effective than doubling the dose of lovastatin alone. As a rule of thumb, LDL cholesterol was reduced by 4% and HDL cholesterol was increased by 8% for each 500-mg increment in niacin ER dose. Lipoprotein (a) levels also decreased with niacin ER doses of 1,000 mg or more.

	Combination treatment was safe and generally well tolerated. Flushing induced by niacin ER was not a major clinical problem. Although 62% of patients experienced flushing, 38% of patients did not. Only 6% withdrew because of flushing over the 5 months of treatment. Adverse events other than flushing were similar among treatment groups.

	The many beneficial effects of niacin ER/lovastatin therapy on lipid profiles  in men and women are clinically applicable to the current evidence-based guidelines for lipid treatment and CAD prevention.3,7 The 2004 update to the National Cholesterol Education Program Adult Treatment Panel III guidelines reemphasized the importance of controlling LDL cholesterol in patients at higher risk, such as those with CAD, diabetes, and metabolic syndrome.7 The update cites new clinical trial data that clearly show for the first time the dependence

	than 100 mg/dL.8-11 Revisions to the guidelines include new optional LDL cholesterol goals of less than

	70 mg/dL for patients at very high risk and less than 100 mg/dL for those at moderately high risk. They also emphasize the routine use of standard doses of statins or combination therapy to lower LDL cholesterol by 30% to 40% or more. For high-risk patients with high triglycerides or low HDL cholesterol levels, or both, the guidelines suggest combination therapy using niacin or a fibrate. Furthermore, the new American Heart Association guidelines for CAD prevention in women have raised the HDL cholesterol treatment goal from 40 mg/dL or greater to

	The broad endorsement of these two guidelines by leading professional groups and experts is setting new national standards of practice. To successfully achieve the increasingly aggressive and often multiple targets, clinicians need to become familiar with the lipid-altering profiles of combination therapies. The surface graphs (figure) show the multi-lipid dose-response effects of niacin ER/lovastatin and provide a practical guide for treating patients with more than one lipid abnormality.6 The graphs can be used to select the optimal dose for attaining LDL cholesterol, HDL cholesterol, triglyceride, and, in selected cases, lipo-protein (a) goals. This study design and efficacy summary by surface graphs can provide a useful standard model for the evaluations of other combinations of lipid drugs, and for orderly comparisons among the drugs.

Niacin extended-release with lovastatin  for hypercholesterolemia

	The diagnosis of heart failure rests on the clinical recognition of its typical symptoms, physical examination findings, and the laboratory evaluation. Other abnormalities, including pulmonary disease, obesity, and renal insufficiency, can mimic many of the clinical findings of heart failure. A reduced left ventricular ejection fraction on echocardiography in a patient with presumed heart failure confirms the presence of a cardiac abnormality and makes it likely that the patient’s signs and symptoms are indeed due to heart failure. Thus, a decreased left ventricular ejection fraction has been required for entry into most of the pivotal trials of therapy for heart failure. Many patients with heart failure, however, do not have a reduced left ventricular ejection fraction. Thus, there has been a search for an objective measure that will allow confirmation of the diagnosis of heart failure.

	B-type natriuretic peptide (BNP) has received the most interest as a potential biomarker for the presence and severity of heart failure. BNP is one of a family of natriuretic peptides. It is predominantly synthesized and released from the ventricular myocardium in response to distention. BNP, and other natriuretic peptides, produce vasodilatation and natriuresis and counteract many of the detrimental effects of neurohormonal activation in heart failure. As such, increased BNP is a compensatory response to heart failure, and BNP is elevated in most patients with heart failure. The elevation of BNP is related to the severity of heart failure, and a marked elevation indicates a poor prognosis. BNP levels may decrease in response to treatment. Thus, BNP has been proposed as a useful measure for the recognition and management of patients with heart failure.

	Not everyone, however, has endorsed the use of BNP in heart failure. Several concerns have been raised. First, heart failure is a clinical diagnosis, and BNP should not be used in isolation to diagnose or exclude its presence. Second, BNP is not completely sensitive or specific. For example, it is not elevated (> 100 pg/mL) in some patients with compensated chronic heart failure and in rare patients with severe, decompensated heart failure. It is possible that in these rare patients, the severe decompensation of heart failure is due to exhaustion of the ventricle’s ability to synthesize the compensatory BNP. Furthermore, BNP can be elevated in other conditions, such as myocardial infarction, pulmonary embolism, and renal impairment.

	Another concern is that BNP levels rise with age in subjects without apparent cardiac disease, especially in women. For example, patients in their eighties who are in the upper quintile of “normal” have a mean BNP level greater than 100 pg/mL; however, these elderly subjects without apparent cardiac disease but with an elevated BNP have a poor prognosis. Thus, it is possible that although they are without apparent cardiac disease, the elevated BNP may indicate a compensatory response to subclinical heart failure.

	The study by Mueller and colleagues (page 32) indicates that BNP measurement is not only useful in recognizing heart failure in patients with acute dyspnea, but it also improves their care by reducing length of hospital stay and associated costs. Thus, this study makes an important contribution to our understanding of the practical and economic advantages of measuring BNP in patients presenting with acute dyspnea.

	This study adds to the growing data supporting the utility of measuring BNP levels for the recognition and management of patients with heart failure. Like any test, it is not perfect; it must be interpreted in light of the entire clinical situation. There is no real conflict between

	the clinical assessment of heart failure and the measurement of BNP. Optimally, clinical judgment is based on an integrated assessment of all the important information about a patient. In the case of possible heart failure, especially when it presents as acute dyspnea, the important information now includes measurement of BNP. When used in this manner, BNP should improve the care of patients being evaluated and treated for heart failure. n

BNP as a biomarker for heart failure 

	The management of patients presenting to the emergency department with acute dyspnea is challenging. The rapid and accurate differentiation of congestive heart failure (CHF) from other causes of acute dyspnea, in particular, often remains elusive. Even after reviewing the patient’s history, physical examination, electrocardiogram, and chest radiograph, the diagnosis may still be uncertain, resulting in misdiagnosis and a delay in instituting appropriate treatement.1-4 Misdiagnosis of heart failure can lead to morbidity, increased time to discharge, and higher treatment costs because treatment methods for patients with other conditions, such as chronic obstructive pulmonary disease, may be detrimental in heart failure patients, and vice versa.1-4

	B-type natriuretic peptide (BNP) is a 32-amino acid polypeptide secreted from both the left and the right cardiac ventricles in response to ventricular volume expansion and pressure overload.5-8 Recent studies have suggested that this neurohormone is reliably elevated in the setting of decompensated CHF and may be very helpful in its diagnosis1,2; however, the real impact of this new diagnostic test on the evaluation and management of patients with dyspnea in clinical practice and whether its use may affect patient outcome or cost of treatment is unknown. We performed the BNP for Acute Shortness of Breath Evaluation (BASEL), a prospective, randomized, controlled, single-blinded study, to test the hypothesis that a BNP-guided diagnostic strategy would improve the evaluation and management of patients presenting to the emergency department with acute dyspnea and, thereby, reduce time to discharge and total cost of treatment.

	The BASEL study was conducted in the emergency department of the University Hospital in Basel, Switzerland, and included 452 patients. According to a computer-generated randomization scheme in a 1:1 ratio without stratification, patients were randomly assigned to receive a diagnostic strategy with (BNP group, n = 225) or without (clinical group, n = 227) the use of BNP levels provided by a rapid bedside assay (figure 1). Patients with severe renal disease (serum creatinine > 250 µmol/L) and patients in cardiogenic shock were excluded.

	The BASEL study recruited a consecutive series of elderly patients with extensive comorbidity. Slightly more than 40% of patients were women. Patient history included coronary artery disease and hypertension in one half of the patients, chronic obstructive pulmonary disease in 30%, any pulmonary disease in 50%, and diabetes in one fourth of the patients. Baseline characteristics were well matched between the study groups.

	The median time interval from presentation to initiation of the appropriate therapy according to the final discharge diagnosis was reduced from 90 minutes in the control group to 63 minutes in the BNP group (P = .029). The use of BNP levels significantly reduced the need for hospital admission or intensive care. Time to discharge was significantly reduced in the BNP group (median 8.0 days) as compared with the

	P = .001). This translated into a significant reduction in mean total treatment cost from $7,264 in the clinical group to $5,410 in the BNP group (P = .006; figure 2). Twenty-one patients (9%) in the clinical group and 13 patients (6%; P = .19) in the BNP group died in-hospital.

	The BASEL study is the first randomized, controlled study of BNP testing in the emergency diagnosis of patients with acute dyspnea. In conjunction with other clinical information, the use of BNP levels reduced the time interval to initiation of the most appropriate therapy, the need for hospital admission and intensive care, and the total treatment time and total cost of treatment. Considering the enormous morbidity associated with dyspnea and the considerable cost associated with CHF, chronic obstructive pulmonary disease, and other causative disorders,1-5 BNP testing seems to represent a major advance in the management of patients with dyspnea.

	The total treatment cost in this study was similar to the current expenses in the United States. In 1997, an estimated $5,501 was spent for each hospital-discharge diagnosis of heart failure.9 BNP testing reduced the total cost of treatment by 26%. Extrapolating this finding to the cost of hospitalizations for heart failure in the United States in 2000 of $24 billion,9 the use of this simple blood test may save up to $7 billion each year in the United States alone.

	The findings of the BASEL study are complementary to those of observational studies in which BNP levels were validated against a retrospectively adjudicated diagnosis of heart failure by independent cardiologists. In the largest of these, the Breathing Not Properly Multinational Study,1,2 the combination of clinical assessment and BNP testing was more accurate than clinical assessment alone. The diagnostic accuracy of BNP at a cutoff of 100 pg/mL was 83%, its sensitivity was 90%, and its specificity 76%. The BASEL study extends these observations and shows that due to this more accurate diagnosis, BNP testing, in fact, improves patient outcome and resource utilization.

	The clinical experience with BNP testing is limited. The interpretation of BNP levels applied in this study was based on the available data at the time the study protocol was prepared (figure 3). Additional studies are needed to help optimize the use of BNP testing in clinical practice. Normal values corrected for age, sex, and renal function may be a significant advance.10,11

	The BASEL study demonstrates that, combined with other clinical information, rapid measurement of BNP improves the management of patients with acute dyspnea in the emergency department, thereby reducing time to discharge and total cost of treatment. n

Natriuretic peptide and the management  of acute dyspnea

Circadian variation  in primary angioplasty for acute MI

	Primary percutaneous coronary intervention (PCI) is the treatment of choice for acute myocardial infarction (MI). This lifesaving procedure, therefore, needs to be available 24 hours a day, 7 days a week. It is important to determine whether the outcome of primary PCI is as good during off-hours as during regular working hours in the hospital. Two important variables may play a role in determining this.

	First, is the pathophysiology of acute MI or the presentation of the disease different during normal working hours compared with off-hours? There is a circadian variation in the time of symptom onset for patients with acute MI as well as with other cardiovascular events, such as unstable angina, sudden cardiac death, transient MI, and ischemic stroke.1-6 Objective assessment with plasma creatine kinase levels has shown a marked circadian variation in the onset of MI, with a peak incidence occurring between 6:00 am and noon and a lower incidence occurring at night.1

	Second, is there a difference in the care delivered during routine working hours compared with off-hours? The Maximal Individual Therapy in Acute Myocardial Infarction (MITRA) study group examined patterns of performance of primary PCI between patients receiving treatment during daytime and nighttime hours.7 Of 491 patients, 77% were treated during the day, and only 23% were treated at night. Patients in the night group had a shorter time to presentation. There was no significant difference between the two groups in clinical outcome. Because of the huge difference between the number of nighttime-treated patients in the eight centers (from 8% to 44%) and the small number of patients treated at night (23%), selection bias may have played a role in the MITRA analysis. The effect on clinical outcome of circadian variations in either symptom onset or treatment has not been studied in hospitals, where all acute ST-segment elevation MI patients are treated with primary angioplasty. We therefore investigated the circadian variation in symptom onset, hospital admission, and first balloon inflation in patients with acute MI. Additionally, we studied the influence of these circadian patterns on clinical outcome in patients with acute ST-segment elevation MI treated with primary angioplasty during routine working hours and during off-hours.

	We studied 1,702 consecutive patients with acute ST-segment elevation MI, treated at our hospital, including those who presented within 6 hours after symptom onset. Patient characteristics were described previously.8 The ambulance crew or staff at one of the 11 referring hospitals located within 94 km from our hospital diagnosed acute MI in 860 patients. All patients from these hospitals who require PCI are referred to our hospital.

	The 860 patients received aspirin (500 mg intravenously [IV]) and heparin (> 5,000 IU IV) before transportation to our hospital. For the remaining 842 patients, the diagnosis of acute MI was established in our hospital, and they received IV aspirin and heparin after admission before going to the catheterization laboratory. No patients received fi-brinolytic therapy or glycoprotein IIb/IIIa blockers before angiography. All primary PCI procedures, during both routine working hours and off-hours, were performed by senior staff members. Follow-up

	for 30 days was completed for all patients.

	The frequency of symptom onset, hospital admission, and first balloon inflation was associated with a circadian variation. The patterns of frequency of symptom onset, hospital admission, and first balloon inflation were similar, with some circadian variation (figure 1). There was also a circadian variation in the 30-day mortality rate. The mortality rate in the first 30 days after successful angioplasty was 2.2%, compared with 17% after failed angioplasty, confirming the strong relationship between mortality and angioplasty failure (figure 1).

	The majority of patients (53%) were treated during the routine working hours of our hospital, between 8:00 am and 6:00 pm. There was no difference in clinical baseline characteristics and antegrade flow in the infarct-related artery between patients treated from 8:00 am to 6:00 pm and those treated from 6:00 pm to 8:00 am. Patients treated between 6:00 pm and 8:00 am had a significantly higher rate of failed angioplasty procedures compared with patients treated between 8:00 am and 6:00 pm (figure 2). The 30-day mortality rate was also higher for patients treated from 6:00 pm to 8:00 am compared with patients treated from 8:00 am to 6:00 pm (figure 3).

	Our study confirms that there is a circadian variation in the time of symptom onset, hospital admission, and treatment of acute MI. Because of the distribution in time of onset of MI, most primary angioplasty procedures are performed during regular working hours, from 8:00 am to 6:00 pm. The observation that patients treated during off-hours had a worse outcome, with a higher rate

	of angioplasty failure and, as a result, a higher 30-day mortality rate, is the more important finding. There may be three reasons for this. First, there may be different baseline clinical characteristics associated with better outcomes for patients who are treated during the day; among patients who have MIs at night, an increased risk of congestive heart failure has been reported.9 In our study, there were no differences in the baseline clinical characteristics, including Killip class, between patients treated during the day and those treated at night, suggesting that this mechanism was not involved in our study.

	Second, the efficacy of reperfusion therapy may be at least partly related to circadian variations. These variations have been documented for platelet aggregation,10 coronary flow, viscosity, cortisol levels, epinephrine levels, and activated partial thromboplastin time and thrombin time. It has been suggested that fibrinolysis increases at night, as shown by levels of tissue-type plasminogen activator and other factors related to natural fibrinolytic activity. Additionally, a circadian pattern in the efficacy of streptokinase and tissue-type plasminogen activator has been reported.11,12 The efficacy of several thrombolytic agents may thus be related to the circadian variation in the balance between prothrombotic mechanisms and the natural fibrinolytic system. The success of primary angioplasty may also be affected by this balance.

	Third, quality of care may be related to the time of day the patient receives treatment. The time of delay from admission to first balloon inflation13 and hospital and physician volume14 are reportedly related to outcome after primary angioplasty. It is therefore conceivable that the quality of care delivered during the day may differ from the care given

	at night because of variations in the performance of the physicians, catheterization laboratory, and coronary care staff. Even high-volume centers may have significant intercenter variation in outcome for patients treated with primary angioplasty.15

	Clinical outcome is influenced by time from hospital admission to first balloon inflation, which may serve as an indication of quality of care. Our study showed no circadian variation in the time from hospital admission to first balloon inflation, suggesting consistent quality of care; our hospital strives to provide optimal quality care for large numbers of patients with acute MI 24 hours a day, 7 days a week. This mechanism, however, appears to at least partly explain the results of our study.

	Whether the higher incidence of failed angioplasty for patients treated during off-hours is related to circadian variations in the balance between natural prothrombotic and fibrinolytic factors or to the quality of care could not be analyzed by our data. It is necessary to develop techniques to evaluate quality of care for patients with acute ST-segment elevation MI who are treated with primary angioplasty, and if the balance between prothrombotic mechanisms and the natural fibrinolytic system

	is shown to play a role, beneficial adjunctive medication should be developed.

	There is a circadian variation in patients presenting with acute MI, which has an enormous effect on

	the success of primary angioplasty. Most patients are treated during regular working hours, and they have a higher procedural success rate and a better clinical outcome than patients treated during off-hours.

	Primary coronary intervention for acute myocardial infarction

	Primary coronary intervention for acute myocardial infarction is rapidly becoming the most preferred way of treating this life-threatening acute event. Boston has a demonstration project1 that has been ongoing for several years in which emergency calls for ST elevation acute myocardial infarction (STEMI) are triaged to hospitals that perform acute coronary intervention around the clock within predetermined time intervals and with predetermined outcomes. New York State is contemplating initiating a similar project. All of this

	me because I have believed in this procedure for acute infarct patients since 1984. It only makes sense,

	however, if outcomes are as good as or better than those obtained with  thrombolytic agents.

	The study by Henriques and colleagues (page 27) from the Netherlands has identified from their large experience with acute infarct intervention that there were circadian differences in outcome in their cohort of acute infarct patients. Is this because there are circadian differences in the patient population, or differences in the performance of the procedure? An analysis of their 1,702 patients treated between 1994 and 2000 showed that 860 patients were diagnosed either in the ambulance or at another hospital, while the rest (842) were diagnosed in the interventional hospital.

	The majority of patients (53%) were treated during normal working hours (8:00 am to 6:00 pm). There were no differences in clinical variables and antegrade flow of the infarct-related artery between the two groups (on-hours and off-hours). There was a significant difference, however, in the number of failed procedures and in the 30-day mortality rate between successful and failed procedures. There was a significant 30-day mortality difference between those patients treated during regular hours (8:00 am to 6:00 pm) and those patients treated after hours (6:00 pm to 8:00 am).

	Although there have been reports of more congestive heart failure in nighttime infarcts,2 there was no difference in Killip class between these two groups. Other circadian differences have been documented in the literature for variables that might influence intervention outcome, such as platelet aggregation, levels of tissue plasminogen activator, etc.3 The literature has also described a circadian variation in thrombolytic therapy.4  The authors verified that door-to-balloon times for both groups were the same.

	In summary, although the authors adequately documented the difference in outcomes between these two groups of patients, they were at a loss, as am I, to adequately explain the reasons for this difference. Naturally, none of us are willing to believe that the quality of care is “different” after regular working hours. n

Primary coronary intervention for acute myocardial infarction

	Case report: Combination therapy for metabolic syndrome

	A 62-year-old man had a history of hypertension and dyslipidemia. He had never smoked, walked regularly, and watched his diet; however, he had a family history of premature myocardial infarction. His medications included hydrochlorothiazide, 50 mg/day; atenolol, 100 mg/day; and simvastatin, 20 mg/day. On physical examination, his blood pressure was 122/76 mm Hg, his height was 5’10”, and his weight was 190 lb (BMI, 27.3). The rest of the physical examination was normal. A fasting lipid profile showed the following values: total cholesterol, 185 mg/dL; low-density lipoprotein (LDL) cholesterol, 112 mg/dL; high-density lipoprotein (HDL) cholesterol, 34 mg/dL; and triglycerides, 178 mg/dL. The lipoprotein (a) level was 32 mg/dL (normal, < 30 mg/dL), and the fasting blood glucose level was normal.

	Like many patients, this man had multiple coronary artery disease (CAD) risk factors and metabolic syndrome as a result of his dyslipidemia, hypertension, and severe overweight. His 10-year Framingham CAD risk was 16%, placing him at moderately high risk. Based on the Adult Treatment Panel (ATP) III of the National Cholesterol Education Program, his LDL cholesterol level was below the minimum goal of less than 130 mg/dL; however, in the recent ATP III update, an optional goal of less than 100 mg/dL has been recommended. He also had many other lipid abnormalities, including low HDL cholesterol, increased triglycerides, and increased lipoprotein (a) levels.

	After normal results were obtained on a stress test, the patient was counseled about the importance of a heart-healthy lifestyle, emphasizing diet, exercise, and weight loss. Successful lifestyle changes can significantly improve multiple risk factors, allowing reduction and possibly withdrawal of some medications in the long term. Doubling the dose of a statin generally results in an additional 6% reduction in LDL cholesterol, with lesser effects on other lipid parameters. To achieve this patient’s composite lipid targets, the simvastatin was discontinued, and niacin ER/lovastatin was gradually titrated to a dose of 2,000/40 mg once daily at bedtime. An aspirin, 30 minutes prior to dosing, was also added to the regimen to limit flushing and for CAD prevention.

	At follow-up, his lipid panel showed good control of all lipid parameters: LDL cholesterol, 94 mg/dL; HDL cholesterol, 42 mg/dL; triglycerides, 145 mg/dL; and lipoprotein (a), 27 mg/dL. The niacin ER/lovastatin was well tolerated, with some initial flushing but no myalgia or increases in fasting blood glucose, creatine kinase, or liver function levels. This case shows the advan-

	tages of combination therapy in controlling multiple lipid abnormalities compared with statin therapy alone in a patient with the metabolic syndrome. n

January 2005

The dangers of nonadherence to cardiovascular medications

Calendar

The fate of the fasting lipid profile

Case report: A patient with cardiogenic shock and impaired fasting glucose

Impaired fasting glucose and cardiogenic shock

Niacin extended-release with lovastatin for hypercholesterolemia

BNP as a biomarker for heart failure

Natriuretic peptide and the management of acute dyspnea

Circadian variation in primary angioplasty for acute MI

Primary coronary intervention for acute myocardial infarction

Case report: Combination therapy for metabolic syndrome

Latest News