Highlights from the American Heart Association Annual Scientific Sessions 2004: November 7-10, 2004

Study 

Cost-Effectiveness of ICD Therapy in the Sudden Cardiac Death in Heart Failure Trial (SCD-HeFT)

Presenter 

Dr. Daniel B. Mark, Duke Clinical Research Institute, Durham, NC

Background 

Implantable defibrillators (ICD)s have been shown to reduce the mortality in patients with congestive heart failure due to ischemic etiology with depressed ejection fraction (EF <30%).1., 2. The Sudden Cardiac Death Heart Failure Trial (SCD-HeFT) evaluated the benefit of implantable defibrillators as primary prevention of death in patients with stable heart failure from both ischemic and non-ischemic causes. The primary results of the trial were presented at the American College of Cardiology meeting in March 2004. This pre-specified, cost-effectiveness analysis was based on the patient data from SCD-HeFT.

Briefly, SCD-HeFT evaluated a single chamber, shock-only implantable defibrillator made by Medtronic in 2521 stable New York Heart Association (NYHA) class II and III heart failure patients with left ventricular systolic EF <35% compared to amiodarone or placebo. Notably, nearly half the patients had heart failure due to non-ischemic causes, 23% of the patients were women, and 23% were minorities. All patients were on optimal medical therapy. The mean follow-up was 41 months. The primary end point was all-cause mortality and the implantable defibrillator reduced this significantly, with a hazard ratio of 0.77.

In order to determine cost-effectiveness, a model was created. Cost data were collected from medical resources, hospital billing data, and outpatient medical costs; statistical imputation was used for missing data. The outpatient medical costs were determined from the 2003 Medical Red Book, and the analysis was performed using 2003 US dollars. The cost was determined to 5 years and lifetime cost-effectiveness was also carried out. The assumptions utilized were ICD generator life of 5 years and a single chamber ICD costing $17,500 (range $13,300-$23,700); extensive sensitivity analyses were performed. The 5-year cost in the placebo arm was determined to be $43,077, and in the ICD arm it was $61,967. Analysis was not carried out in the amiodarone arm because that therapy was not effective.

Results 

The placebo-arm life expectancy was determined to be 8.4 years. For the base case assumptions, ICD therapy resulted in a cost of $33,192 per life year saved. The subgroups had similar costs: EF <30%, $33,509; EF >30%, $29,275; ischemic etiology, $33,603; non-ischemic etiology, $32,170; age <65 years, $39,469; age >65 years, $29,164. If the time to generator change-out was decreased and the upper limit of ICD cost was used, the cost was $53,459. Finally, if the benefit of ICD therapy was assumed to end at 5 years, the cost was $77,175.

Interpretation 

The results of this cost-effectiveness analysis should be put in perspective of currently accepted medical therapies. “Economically attractive” therapy from a societal perspective can be benchmarked with federal reimbursement levels for dialysis in patients with chronic kidney disease, which is approximately $50,000 per year of life gained. In this context, single-chamber, shock-only ICD therapy in patients with heart failure and depressed left ventricular systolic function (EF <35%) is cost effective per year of life gained.

Study 

African-American Heart Failure Trial (A-HeFT)

Presenters 

Dr. Jay N. Cohn and Dr. Anne L. Taylor, Minnesota Medical School, Minneapolis, Minn

Background 

Researchers have long suspected that vasodilator therapy is an important therapeutic intervention in patients with congestive heart failure (CHF). In the mid 1980s, the Veterans Affairs Vasodilator-Heart Failure Trial (V-HeFT) revealed increased survival in CHF patients treated with combination hydralazine/nitrates compared with placebo.3 More recently, however, an analysis of the V-HeFT trials found that although enalapril therapy compared with combination hydralazine/isosorbide dinitrate was associated with decreased mortality among white patients, there was no benefit in black patients.4 Isosorbide dinitrate (ISDN) is a nitric oxide (NO) donor and hydralazine is an anti-oxidant that inhibits NO destruction. These observations led to the hypothesis that there exist racial differences in neurohormonal stimulation and NO effects in patients with CHF, prompting the African-American Heart Failure Trial (A-HeFT), the first heart failure trial conducted only in black patients. A-HeFT sought to test the hypothesis that NO-enhancing therapy with a fixed-dose combination of hydralazine and ISDN improves outcomes in black patients with moderate to severe heart failure.

A-HeFT randomized patients from 161 sites from June 2001 to July 2004, at which time the Data Safety and Monitoring Board terminated the trial prematurely due to evidence of decreased mortality in patients receiving combination hydralazine/ISDN therapy. Criteria for enrollment included self-identification as black; stable New York Heart Association (NYHA) Class III-IV CHF; standard background therapy (with a minimum of 3 months of β-blocker therapy); and left ventricular ejection fraction (LVEF) <35% (or <45% with LV internal end-diastolic diameter >6.5 cm). Patients with uncontrolled hypertension and recent myocardial infarction (MI) or stroke were excluded. Patients were randomized to a fixed-dose combination pill of ISDN (20 mg) and hydralazine (37.5 mg) or placebo, in addition to background therapy for CHF (as determined appropriate by their physicians). The primary efficacy end point was a composite score comprising weighted values for death from any cause; first hospitalization for CHF; and change in quality-of-life (QOL) at 6 months. Secondary end points included individual components of the primary composite score, death from cardiovascular causes, and a number of other CHF measures. The primary analysis was an intention-to-treat analysis with worst case score for missing data. The Minnesota Living with Heart Failure questionnaire was used to assess QOL measures.

Results 

Results of the A-HeFT trial were recently published.5 One thousand fifty patients (out of a planned 1100) were randomized (518 to ISDN/hydralazine, 532 to placebo). No patients were lost to follow-up. There were no significant differences in baseline characteristics between the 2 arms, except for a slightly smaller percentage of men and a slightly higher percentage of diabetic patients in the ISDN/hydralazine group. Of note, ischemic heart failure accounted for only 23% of patients, with hypertensive and idiopathic heart failure accounting for the majority of causes. The mean LVEF was 24%, and 40% of participants were female. Patients were well managed with background medical therapy (70% on angiotensin-converting enzyme inhibitors; 74% on β-blockers; 40% on spironolactone; 60% on digitalis; and 90% on diuretics). At trial termination, the primary composite score was −0.1 ± 1.9 in the group treated with ISDN/hydralazine, compared with −0.5 ± 2.0 in the placebo group (P = .01), consistent with a beneficial effect of ISDN/hydralazine. There were significant differences between placebo and treatment arms for the individual components of the composite score: death 10.2% versus 6.2% (P = .02); first CHF hospitalization 24.4% versus 16.4% (P = .001); change in QOL −2.7 versus −5.6 (P = .02, with greater improvement in QOL in the treatment arm). There was a significant 43% improvement in survival (hazard ratio 0.57, P = .01), and Kaplan-Meier survival curves revealed divergence between the arms at approximately 180 days of therapy, with continued widening thereafter (P = .01 by log-rank test). Side effects of headache and dizziness occurred more frequently in the group treated with ISDN/hydralazine, whereas CHF exacerbations were significantly more frequent in the placebo-treated patients.

Interpretation 

The A-HeFT trial revealed a 43% increase in survival and a 33% decrease in the incidence of first heart failure hospitalization in black patients treated with a fixed-dose combination pill of ISDN and hydralazine compared with placebo. The results also showed a significant improvement in QOL in patients treated with the fixed-dose combination. The investigators conclude that NO-enhancing therapy is an effective adjunctive treatment to standard background CHF therapy for patients with severe heart failure, especially in black patients.

The A-HeFT trial is a noteworthy, well designed trial conducted in a high-risk minority population. The population was a well matched, sick heart failure population with a high prevalence of comorbidities. The 100% follow-up rate is particularly remarkable, as is the high use of background cardiovascular pharmacotherapies and the high proportion of women enrolled (40%, representing the highest percentage of women enrolled to date in heart failure trials). Of note, the leading cause of heart failure in this population was hypertension, raising the question that the benefit of the combination therapy is related to its anti-hypertensive effect and not to NO-related effects. These results must also be weighed in light of the high occurrence of side effects (headache and dizziness), raising concerns about compliance outside of a clinical trial. Regardless, the results are compelling for the inclusion of ISDN and hydralazine in the management of heart failure patients, particularly black patients. This trial highlights the potential for improving outcomes and QOL in minority populations. Further studies will be necessary to unravel the etiology of the benefit seen.

Study 

Prevention of Events with the Angiotensin Converting Enzyme Inhibition (PEACE) Trial

Presenter 

Dr. Marc A. Pfeffer, Brigham and Women's Hospital, Boston, Mass

Background 

Blockade of the renin-angiotensin system with angiotensin-converting enzyme (ACE) inhibitors improves survival in patients with congestive heart failure and depressed left ventricular systolic function.6., 7., 8. This mortality benefit is also present in patients with acute myocardial infarction (AMI).9., 10., 11. Recently, studies have demonstrated a benefit with ACE-inhibitor therapy in patients at high risk for coronary events. The HOPE study with ramipril and the EUROPA study with perindopril found that patients with coronary or other vascular disease or patients with diabetes and other cardiovascular risk factors had a reduction in death and myocardial infarction when treated with ACE-inhibitor therapy.12., 13. Although these trials enrolled lower-risk patients than previous heart failure and acute myocardial infarction trials, patients were still at significant risk for cardiovascular events. Therefore, the goal of the Prevention of Events with Angiotensin Converting Enzyme Inhibition (PEACE) trial was to determine if ACE-inhibitor therapy with trandolapril in addition to routine proven care would reduce death from cardiovascular causes, myocardial infarction, or revascularization in low-risk patients with stable coronary artery disease and normal or slightly reduced left ventricular ejection fraction (EF).

The PEACE trial was a randomized, double-blinded, placebo-controlled trial. The inclusion criteria were age >50 years, coronary artery disease defined as acute MI (AMI), bypass surgery or percutaneous intervention (PCI) at least 3 months before enrollment or >50% stenosis on coronary angiography, left ventricular ejection fraction >40%, and ability to tolerate and comply (≥80%) with medication during a 2-week run-in phase. The exclusion criteria were patients on an angiotensin-blocking agent, any revascularization within 3 months, unstable angina, planned revascularization, pregnancy, creatinine >2.0 mg/dL, or potassium >5.5 mmol/L. The study had 90% power to detect an 18% reduction in the primary end point of death from cardiovascular causes, MI, or revascularization, assuming a 19% incidence in the placebo group.

Results 

A total of 8290 patients were randomized, 4158 to trandolapril 4 mg target dose and 4132 to placebo, between November 1996 and June 2000, with a median follow-up of 4.8 years. The baseline characteristics between the groups were similar. The mean age was 64 ± 8 years, 92% were white, 18% in the trandolapril and 16% in the placebo group had diabetes, 54% to 56% had a prior MI, the mean EF was 58%, and 15% of the patients had an EF >40% but <50%. The mean blood pressure was 133 ± 17/78 ± 10 mm Hg. Use of proven therapies was also similar, with 90% use of aspirin, 70% use of lipid-lowering agents, and 60% use of β-blockers in the population.

The primary end point of death from cardiovascular causes, MI, or revascularization occurred in 21.9% of the trandolipril group compared with 22.5% of the placebo group (hazard ratio for the trandolipril group 0.96, 95% CI 0.88-1.06, P = .43). This effect was similar across all subgroups. The secondary end points of death from cardiovascular causes, all-cause death, non-fatal MI, or revascularization either by PCI or bypass were all similar between the 2 groups. Of note, after 36 months, placebo led to a 1.4/2.4 mm Hg blood pressure change compared to a 4.4/3.6 mm Hg change in the trandolapril group (P < .001).

Interpretation 

In stable patients with coronary artery disease without significant left ventricular dysfunction on current medial therapy, the addition of ACE-inhibitor therapy with trandolapril does not reduce death from cardiovascular causes, MI, or revascularization. The findings of this study demonstrate that in a low-risk coronary artery disease population, mostly without diabetes or heart failure, ACE-inhibitor therapy in addition to standard care is not beneficial.

Study 

ESCAPE (Evaluation Study of Congestive Heart Failure and Pulmonary Catheterization Effectiveness)

Presenters 

Dr. Monica Shah, Columbia University, New York, NY, and Dr. Lynne W. Stevenson, Brigham & Women's Hospital, Boston, Mass

Background 

Pulmonary artery catheterization (PAC) has been used to guide treatment decisions in patients with congestive heart failure (CHF) for a long time; however, the use of this strategy is not clearly supported in the literature. Some non-randomized studies have shown that the use of PAC increases morbidity and mortality in critically ill patients.14 On the other hand, there is evidence that the routine use of PAC can improve symptoms and quality of life (QOL) and decrease hospitalizations.15

The ESCAPE trial's primary hypothesis was that therapy guided by clinical assessment and PAC in patients with severe CHF leads to fewer days dead or hospitalized over a period of 6 months than therapy guided by clinical assessment alone. Major inclusion criteria were decompensated HF, history of advanced HF despite aggressive medical therapy, left ventricular ejection fraction (LVEF) <30%, systolic blood pressure (SBP) <125 mm Hg, symptoms of HF for >3 months, therapy with ACE inhibitors and diuretics, and unlikely transplant within 6 months. Exclusion criteria were acute decompensated HF requiring urgent PAC, serum creatinine >3 mg/dL, need of mechanical support, and use of inotropic drugs. The primary end point was the number of days patients were not dead or hospitalized over a period of 6 months. Secondary end points were serum brain natriuretic peptide (BNP) levels, peak oxygen consumption (VO2 max), 6-minute walk test, and QOL end points. Over a period of 4 years, eligible patients hospitalized in 26 institutions were randomized in a 1:1 fashion to receive therapy guided by clinical assessment and catheterization (PAC arm) or clinical assessment alone (CLIN arm). Treatment goals were resolution of signs and symptoms of CHF (for both arms), and pulmonary capillary wedge pressure (PCWP) ≤15 mm Hg and right atrial pressure (RAP) ≤8 mm Hg for the PAC arm.

Results 

From a total of 433 patients, 218 were randomized to the PAC arm and 215 to the CLIN arm. There were no significant differences in baseline characteristics. The rate of adverse events was 17% (overall), 22% (PAC), and 11% (CLIN). PAC-related complications were higher in the PAC group (4.2% vs 0.5%). There were no deaths related to PAC. Early mortality was 4.7% and 5% in PAC and CLIN arms, respectively. There were no differences between PAC and CLIN arms in days to death or re-hospitalization (13 vs 13); days to death (45 vs 38); number of re-hospitalizations per patient (2 vs 2); and days in hospital (11 vs 11). On secondary end points there were no difference in BNP levels, VO2 max, and 6-minute walk. Time trade-off (TTO), a QOL measure, was better in patients assigned to the PAC group.

Interpretation 

The ESCAPE trial showed that the use of PAC in patients with CHF is safe but does not affect clinical outcome. These results are consistent with other trials in ICU patients. Based on this result, despite the fact that the effect of PAC seems to be neutral, there is no evidence to support its routine use in this clinical setting. Although all of the patients in this trial were in NYHA class IV, we cannot extrapolate these results to patients who need urgent catheterization. Furthermore, specific regimens to achieve therapeutic goals were not specified in the study design, which can be a limitation. The fact that patients in the PAC group showed a better QOL index is a very interesting finding and should be explored in futures studies. Information about echocardiographic and functional analyses will be provided by the trial investigators in the future. Ultimately, physicians should be aware that PAC-guided therapy is safe but does not improve outcome.

Study 

Clinical Evaluation of the CorCap Cardiac Support Device in Patients with Dilated Cardiomyopathy

Presenter 

Dr. Douglas L. Mann, Baylor College of Medicine, Houston, Tex

Background 

In congestive heart failure (CHF), the progression to left ventricular (LV) dilatation from an ellipse to a sphere can predict mortality.9., 16. The CorCap cardiac support device (CSD) is a knitted polyester sock that is drawn up and anchored over the ventricles, designed to address this dilation. It is the first investigational device to test whether halting remodeling can alter the unfavorable natural history of CHF. Early animal and human studies have shown that the CSD is safe and is associated with improvements in LV structure and clinical symptoms of CHF. Importantly, constrictive physiology and inhibition of coronary flow have not been observed. The investigators hypothesized that the CSD would prevent LV remodeling and reduce LV dilation.

The design of the trial was a prospective, multi-center, randomized clinical trial enrolling 300 patients with NYHA class II-IV CHF, left ventricular end-diastolic diameter (LVEDd) >60 mm, EF <35%, and 6-minute walk test <450 feet. The majority of patients were Class III CHF (81%) with idiopathic cardiomyopathy (CM) (61%). Ischemic CM comprised only 10%. One hundred ninety-three patients undergoing mitral valve repair/replacement (MVR) were randomized to MVR alone (102) or MVR plus CSD (91). The remaining 107 patients were randomized to either medical therapy alone (50) or to the CSD (57).

The primary end point was a 12-month composite of death, change in NYHA class, and the need for major cardiac procedures (indicating worsening CHF). For the composite end point, patients were described as improved, unchanged, or worsened.

Results 

Patients randomized to the CSD demonstrated more improvement (38% vs 27%) and less worsening (37% vs 45%) with an odds ratio (OR) of 1.73 (P = .02) in favor of the CSD group. The major driver of the composite end point was the need for major cardiac procedures (19% vs 33% in CSD vs control, P = .007). Mortality and change in NYHA class were not significantly affected.

The CSD group had a significant reduction in LVEDd (P = .009), LVEDs (P = .017), and sphericity index (P = .026), which improved at each measured time point in the 12 months of follow-up. The LV ejection fraction was not significantly improved (P = .45), but there was a trend to improvement over the 12 months. Patient quality of life measures including the Minnesota Living with Heart Failure Index and SF36 were both improved with the CSD (P = .05 and P = .015, respectively). However, re-hospitalization was not decreased (P = .44).

The group that only received the CSD without MVR favored the device with an OR of 2.56 (P = .03). Additionally, the control group with only MVR benefited in comparison to medical therapy alone. Adverse events were non-significant.

Interpretation 

The study concluded that symptomatic heart failure patients (mostly non-ischemic Class III CHF), despite optimal medical therapy, benefit from the CorCap device. Additionally, two-thirds of the patients received an MVR. Because medical therapy, cardiac resynchronization, and MVR do not help everyone, this device will be synergistic with and additive to existing therapies.

The trial limitations include the small sample size and the low inclusion of ischemic CM. In spite of the small size, however, it is the largest randomized surgical trial in heart failure ever conducted. Additionally, despite signs that patients feel better and overall LV parameters were improved, there was no impact on CHF class, mortality, or hospitalizations. However, the LVEDd improved with the CorCap more than has been reported by ACE-inhibitors,17 β-blockers,18 and cardiac resynchronization therapy.19 Furthermore, the NYHA class improvement with the CSD may have been underestimated because the NYHA class assessment in the control arm was censored for major cardiac procedures and death. In the future, developing minimally invasive placement of the CorCap may increase its application.

Study 

Impact of a Simple Regimen of Low Molecular Weight Heparin (Reviparin) in Preventing Mortality, Reinfarction, and Strokes in over 15,500 Patients with ST-Elevation Acute Myocardial Infarction: The CREATE Trial

Presenters 

Dr. Denis Xavier, Bangalore, India, and Dr. Salim Yusuf, Hamilton, Ontario, Canada

Background 

Low-molecular-weight heparin and unfractionated heparin have been used in conjunction with standard antiplatelet and reperfusion therapy in patients with acute ST-elevation myocardial infarction (STEMI), despite the absence of data demonstrating a mortality benefit for heparin compounds in this setting. The Clinical Trial of Metabolic Modulation in Acute Myocardial Infarction Treatment Evaluation (CREATE) investigators therefore studied whether reviparin administered subcutaneously could reduce cardiac events and mortality in >15,000 STEMI patients.

Methods 

A total of 15,570 patients from India and China admitted for acute STEMI were randomized to receive standard care plus subcutaneous reviparin dosed by weight twice daily for 7 days versus standard care plus placebo subcutaneously twice daily for 7 days. These patients also were randomized in a 2×2 factorial design to glucose-insulin-potassium (GIK) therapy plus standard care or no GIK infusion plus standard care as part of the CREATE-ECLA trial. The primary outcome was a composite of total mortality, myocardial infarction, and strokes at both 7 and 30 days.

Results 

Reviparin reduced the incidence of the primary composite outcome (total death/reinfarction/strokes), total mortality alone, and reinfarctions alone at both 7 and 30 days (Table I). The stroke rate did not differ significantly at 30 days (80 strokes in the reviparin arm, 64 in the placebo arm, P = .193). Intracranial hemorrhage and fatal bleeding were included in the primary end point (as strokes and deaths, respectively); in addition, there was a slight increase in other life-threatening bleeds (17 [0.2%] vs 7 [0.1%], P = .07). There was a significant interaction between the efficacy of reviparin and the time to its administration: the hazard ratios for the primary composite end point for administration within 2 hours, 2 to 4 hours, 4 to 8 hours, or >8 hours were 0.70, 0.81, 0.85, and 1.06, respectively (P-value for trend .041).

Table I.

	Outcome (%)	Reviparin	Placebo	Hazard ratio and 95% CIs	P
	Total mortality/reinfarctions/strokes at 7 days (primary)	9.6	10.9	0.87 (0.79-0.96)	.006
	Total mortality at 7 days	8.0	8.9	0.89 (0.80-0.99)	.036
	Reinfarction at 7 days	1.6	2.1	0.75 (0.60-0.95)	.018
	Total mortality/reinfarctions/strokes at 30 days (co-primary)	11.8	13.6	0.87 (0.79-0.95)	.001
	Total mortality at 30 days	9.8	11.3	0.87 (0.79-0.86)	.005
	Reinfarction at 30 days	2.0	2.6	0.77 (0.62-0.95)	.014

Interpretation 

Reviparin reduced total mortality, reinfarction, and the composite of death/reinfarction/stroke in patients with STEMI. The investigators reported that treatment of 1000 patients with reviparin would prevent 17 composite events (and 15 deaths) at the cost of 1 additional life-threatening bleed, and concluded that the benefits of reviparin in STEMI clearly outweigh the risks. The interaction between reviparin treatment and the time to its administration is noteworthy and warrants further study. Even though the rates of use of other evidence-based medications in the trial (thrombolytics, aspirin, β-blockers, statins, ACE inhibitors) were excellent, whether Western countries will incorporate the use of low-molecular-weight heparin into standard practice remains to be seen. Although almost 80% of patients underwent reperfusion therapy, only 6% of the total population underwent PCI (the rates of which are much greater in Western countries), and 73% underwent fibrinolysis, mostly with streptokinase (compared with the predominant use of tenecteplase or alteplase in Western countries). In addition, the benefit of reviparin was achieved by continuous inpatient administration for 7 days, whereas hospital stays are much shorter in North America. Finally, the cost of therapy is significantly less in India and China than in the United States, and reviparin has not yet been approved by the United States Food and Drug Administration. Nonetheless, this trial is likely to change practice and reduce cardiovascular events on a global scale, and the investigators are to be congratulated for conducting such a large, high-quality, high-impact trial in a low-cost fashion with >99.95% follow-up.

Study 

Impact of Glucose-Insulin-Potassium on Mortality and Morbidity in Over 20,000 Patients with Acute Myocardial Infarction: The CREATE-ECLA International Trial

Presenters 

Dr. Rafael Diaz, Rosario, Argentina, and Dr. Shamir R. Mehta, Hamilton, Ontario, Canada

Background 

Since it was proposed over 40 years ago that an infusion of a glucose-insulin-potassium (GIK) solution may reduce the incidence of ventricular arrhythmias following acute myocardial infarction,20 several small clinical trials have yielded inconsistent effects of GIK therapy in this setting. The Clinical Trial of Metabolic Modulation in Acute Myocardial Infarction Treatment Evaluation-Estudios Cardiológicos Latinoamérica (CREATE-ECLA) international trial was designed to determine the efficacy of GIK in >20,000 acute ST-elevation myocardial infarction (STEMI) patients.

A total of 20,201 patients admitted for acute STEMI from India, China, South America, Pakistan, North America, Europe, and the Middle East were randomized to receive standard care or standard care plus GIK infusion within 12 hours of hospitalization. A “high-dose” GIK regimen consisting of 25% glucose, 50 units/L of insulin, and 80 mEq/L KCl was infused for 24 hours in patients randomized to receive GIK therapy. The primary outcome was total mortality at 30 days; secondary outcomes were reinfarction, nonfatal cardiac arrest, and cardiogenic shock at 30 days.

Results 

The mean age of the population was 58.6 years and 77.8% were male. Rates of use of proven therapies for STEMI were high: 74% received thrombolysis, 9% primary percutaneous coronary intervention (PCI), 97% aspirin, 70% β-blockers, 68% statins, and 72% ACE inhibitors. Follow-up was >99.95% complete. There was no difference between groups in any of the primary or secondary outcomes at 30 days (Table II). Patients in the GIK arm had a significant reduction in recurrent ischemia at 7 days (5.6% vs 6.5%, P = .004), which was sustained at 30 days (P = .036). The GIK arm observed higher rates of hyperkalemia, defined as K >5.5 mEq/L (4.3% vs 1.6%, P < .0001) and hypoglycemic episodes (0.4% vs 0.1%, P = .0006); there were no differences in the rates of pulmonary edema (2.1% for GIK vs 1.9% for standard care, P = .42) or congestive heart failure (17.4% in both arms, P = .88). The baseline glucose was 162 mg/dL (9.0 mmol/L); glucose levels were higher in the GIK arm than in the control arm at both 6 hours (187 mg/dL [10.4 mmol/L] vs 148 mg/dL [8.7 mmol/L]) and 24 hours (155 mg/dL [8.6 mmol/L] vs 135 mg/dL [7.5 mmol/L]).

Table II.

	Outcome at 30 days (%)	GIK (n = 10,088)	Standard care (n = 10,107)	Hazard ratio and 95% CIs	P
	Total death (primary)	10.0	9.7	1.03 (0.95-1.13)	.45
	Cardiac arrest	1.4	1.5	0.93 (0.74-1.17)	.51
	Cardiogenic shock	6.6	6.3	1.05 (0.94-1.17)	.38
	Reinfarction	2.3	2.4	0.98 (0.62-1.17)	.81

Interpretation 

This large trial definitively showed that GIK therapy had no impact on survival or other hard clinical end points in acute STEMI. The results were consistent across all prespecified subgroups, including those who did and did not receive reperfusion therapy. A post-hoc analysis suggested benefit among the subset of patients undergoing primary PCI or receiving alteplase; however, the authors note this should be interpreted with caution as the analysis was data-derived and not prespecified.

Focus should now shift to testing a strategy of achieving normoglycemia in acute myocardial infarction patients. In a landmark randomized, controlled trial of surgical intensive care unit (ICU) patients (mostly post-cardiac surgery), intensive insulin therapy administered with a goal of achieving normoglycemia (glucose 80-110 mg/dL) yielded an impressive 42% mortality reduction in the ICU and 34% reduction in in-hospital mortality compared with standard care (insulin infusion only if plasma glucose >200 mg/dL).21 Even though this strategy has not been replicated in an acute coronary syndrome population, the American Heart Association/American College of Cardiology guidelines state that the achievement of normoglycemia in patients with acute STEMI with complicated hospital courses is a class I recommendation, and a class IIa recommendation for patients with hyperglycemia, even with uncomplicated hospital courses.21 Randomized controlled trials of tight glycemic control are required to substantiate these recommendations.

Finally, it is noteworthy that the CREATE-ECLA investigators enrolled most patients from low- or middle-income countries and conducted this study with almost no support from industry. Rates of evidence-based therapies were high, and follow-up was >99.95% complete. Dr. Mehta noted that the international investigators participated “in the true spirit of scientific camaraderie” and received little to no compensation for their efforts.

Study 

The Pediatric Heart Network Fontan Cross-Sectional Study:

Presenters 

Dr. Bernard J. Clark III, Children's Hospital of Philadelphia, Pa; Dr. Brian W. McCrindle, Hospital for Sick Children, Toronto, Ontario, Canada; and Dr. Andrew D. Blaufox, Medical University of South Carolina, Charleston, SC

Background 

The lack of applicable outcome measures has impeded the development of clinical trials in congenital heart disease patients. In an effort to recruit and follow patients with congenital heart disease, the National Heart, Lung, and Blood Institute established and funded the Pediatric Heart Network (PHN) in 2001. The PHN comprises 7 clinical centers, an independent network chair and a data coordinating center and is currently leading 6 clinical studies. The first study completed was the Relationship Between Functional Health Status and Laboratory Parameters of Ventricular Performance After the Fontan Procedure. The purpose of this multi-center, prospective, cross-sectional study was to determine the interrelationships among health status and measures of cardiac performance in children aged 6 to 18 years with congenital heart disease who have undergone a Fontan procedure as surgical treatment for functional single ventricle. The goal is to quantify the correlation between measures of health status and ventricular performance and develop a data set that will permit identification of a clinically relevant end point for subsequent trials of medical management of the Fontan patient.

A total of 546 subjects were consented for participation in this trial. Five hundred thirty-seven parents completed the Child Health Questionnaire (CHQ),22 and these subjects were included in the first 2 analyses discussed. The median age of the subjects was 11.0 + 3.4 years. There was a male predominance (60%). The cardiac diagnoses varied widely; however, each subject had a functional single ventricle, and there was a slight left ventricular dominance (52%). The majority of Fontan procedures were the intracardiac lateral tunnel type.

The third analysis included the 522 patients who had an electrocardiogram (ECG) and/or exercise testing completed. Any patient who had undergone a Fontan conversion was excluded from this analysis. Any patient taking a β-blocker or a class III antiarrhythmic was excluded from the heart rate analysis only, and any patients who were paced during the resting ECG were excluded from the resting ECG analysis only.

Children are at risk of developing heart failure and other medical complications related to the Fontan procedure.23 The first analysis sought to identify quantifiable measures of ventricular performance which correlate with functional and psychosocial health status in patients following Fontan procedure. Fontan patients may be especially predisposed to poor health-related quality-of-life (HRQOL).24 The second analysis sought to define aspects of HRQOL in Fontan patients. Sinus node dysfunction and atrial tachycardia are common following the Fontan procedure. The third analysis sought to delineate factors associated with heart rate (HR) and rhythm abnormalities in children who have undergone the Fontan procedure and to use these objective criteria to determine how these abnormalities impact the daily lives of these children.

Results 

Seven United States centers participated in this study, a total of 1078 charts were reviewed, and 637 patients were eligible for participation. Of these, 546 consented (86%), and the parents of 537 subjects completed the CHQ. Subjects underwent measurement of brain natriuretic peptide (BNP) (n = 511), exercise testing (n = 415), and cardiac MRI (n = 193). The mean CHQ physical summary score was one standard deviation (SD) below normal (45.3 ± 11.9 vs 53.0 + 8.8, P < .001). Mean CHQ psychosocial summary score was one-half SD below normal (47.2 ± 10.8 vs 51.2 ± 9.1, P < .0001). Peak exercise performance averaged 64% predicted using age- and sex-matched controls (mean peak V02 26.4 ± 6.9 mL/kg/min). Mean ventricular mass-to-volume ratio was 0.88 ± 0.31. Compared to pre-adolescents, post-adolescent subjects had significantly lower exercise performance, higher mass/volume ratio and higher BNP levels. The CHQ physical summary score was positively correlated with peak V02 (r = 0.28, P < .001) and negatively correlated to ventricular mass-to-volume ratio (r = −0.16, P = .053). BNP levels correlated negatively to exercise performance (r = −0.14, P = .01).

The second analysis focused on the relationship of medical and patient characteristics to HRQOL in adolescents following the Fontan procedure. The medical experience for these patients ranged widely and, for some, was quite intensive. Patients underwent a mean of 3 cardiac surgeries before their Fontan procedure. The mean length of post-operative stay at Fontan procedure was 18 days (range 3-278 days). New onset of complications during follow-up included arrhythmia (21%), ventricular dysfunction (11%), thrombosis (8%), protein-losing enteropathy (4%), and stroke (2%).

The patients scored significantly lower than normal in several CHQ domains, including physical functioning, impact on roles by both emotional problems and physical limitations, mental health, self-esteem, general health, and impact on the parents both in terms of emotional distress and time. The mean physical functioning and psychosocial summary scores for Fontan patients were lower than those for normal children but also lower than for other cardiac conditions, including arterial switch patients25 and transplant patients, as well as other chronic illnesses, such as diabetes and obesity. Fontan patients had other co-morbidities including allergies (16%), asthma (14%), other chronic respiratory ailments (11%), and orthopedic problems (11%). Sensory deficits were also more common than normal, with 33% having visual problems, 26% speech problems, and 7% deafness. Psychological and behavioral difficulties were particularly more prevalent in these patients, with 46% having attention problems, 42% learning problems, 23% behavioral problems, 17% problems with anxiety and 8% with depression. Multivariate regression modeling identified medical variables associated with poorer physical functioning summary scores. These variables included greater weight at Fontan procedure (P = .01), lack of fenestration performed during Fontan (P < .001), concurrent surgical procedures performed during Fontan (P < .01), arrhythmias occurring during follow-up (P = .02), and a greater number of current medications at study enrollment (P < .001). Non-cardiac variables were also associated with poorer physical functioning summary scores including asthma, non-asthma respiratory problems, orthopedic problems, learning problems (P < .001, all) and lower socio-economic status (SES) (P = .04). A similar multivariable model determined medical history factors associated with poorer psychosocial functioning summary scores, including the presence of behavior and learning problems (P < .001), anxiety (P < .001), depression (P = .02) and lower SES (P < .001).

The third analysis examined factors associated with HR and rhythm abnormalities in children who have undergone the Fontan procedure to determine how these abnormalities impact the daily lives of these children. Seventy percent of patients were in sinus rhythm on the resting ECG; 20% had a superior P-wave axis; 14% had a current pacemaker; and nearly 10% had a history of atrial tachycardia. The average resting HR for the group was 76 ± 16 beats/min. Nearly 27% of patients had resting heart rates below the 5th percentile for age. The average peak HR during exercise was 155 ± 23 beats/min, which was 74% of that predicted. Overall, the group had an abnormal HR response to exercise, with nearly 44% of subjects with peak HR <75% of their predicted peak HR. Patients with abnormally low peak HR with exercise were younger (P < .001) and had a more superior P-wave axis (P = .01) than did those with normal peak HR. Patients with atrial tachycardia were less likely to be in sinus rhythm on resting ECG, less likely to have a normal P-wave axis, more likely to have undergone an atrial pulmonary connection type of Fontan, and more likely to currently have a pacemaker. Correlations with functional outcome revealed that a higher resting HR was associated with a lower oxygen consumption at anaerobic threshold (VAT) (r = −0.19, P = .002) and worse functional outcome (r = −0.13, P = .007). Additionally, a higher peak HR was associated with a better functional outcome (r = 0.13, P = .012). A history of atrial tachycardia was associated with a significantly lower CHQ physical score (P < .001) and a trend toward lower VAT (P = .08). Additionally, patients who currently had a pacemaker had a significantly lower CHQ physical score (P = .010).

Interpretation 

Parents' perceptions of their children's physical and psychosocial functioning following the Fontan procedure were decreased compared to controls. Exercise performance was significantly lower in children following the Fontan compared to healthy children. Better exercise performance was associated with higher CHQ scores. Older children had significantly lower exercise performance and higher ventricular mass/volume ratio and BNP levels. Cardiac MRI findings and BNP levels had only a weak correlation with other variables, and it may be difficult to use these as surrogate end points.

There are important deficits in many aspects of physical and psychosocial HRQOL in children and adolescents after Fontan procedure. For physical functioning, medical history had some influence, and current non-cardiac conditions and SES were also important determinants. For psychosocial functioning, medical history had little direct influence, while current psychosocial morbidities and SES were important determinants. Strategies to address poor HRQOL in Fontan patients might emphasize coordinated and effective management of non-cardiac medical conditions, and prevention, identification, and management of learning, behavioral, and emotional problems, and specific targeting of interventions directed at patients from low-income households.

Children who have undergone the Fontan procedure have a high incidence of non-sinus rhythm, low resting and peak heart rates, and a high incidence of pacemaker placement and atrial tachycardia. High resting HR, low peak HR, history of atrial tachycardia or having a pacemaker are associated with poor functional outcome. In conclusion, HR and rhythm abnormalities are common in children who have undergone the Fontan procedure and these abnormalities may impact their daily lives.

Study 

GEMINI (Metabolic Effects of Carvedilol vs Metoprolol in Patients with Type 2 Diabetes Mellitus and Hypertension)

Presenter 

Dr. George L. Bakris, Rush University Medical Center, Chicago, Ill

Background 

Patients with diabetes mellitus with poor glycemic control have more adverse cardiovascular events than diabetic patients with good glycemic control.26 While blocking the renin-angiotensin system (RAS) has become the cornerstone of antihypertensive therapy in diabetics due to improved cardiovascular event rates, renal protectio, and improved glycemic control, different classes of β-blockers have never been studied with regard to their antihypertensive effects and glycemic control in patients with diabetes. Previous studies have implicated β-blocker causality in increasing fasting glucose by 28 mg/dL and HbA1c by 1%.27

The GEMINI investigators' central hypothesis was that for hypertensive control, carvedilol (a β-blocker with a known favorable glycemic profile) would maintain better glycemic control than metoprolol tartrate (a β-blocker with known insulin resistance characteristics). The GEMINI trial was a randomized, double-blind, parallel group, multi-center design that randomized 1235 diabetic patients to either carvedilol or metoprolol tartrate. Patients were already on RAS blockers and were enrolled if their blood pressure (BP) was between 140 and 179 systolic or 90 and 109 diastolic. Doses were escalated from 6.25 mg twice daily of carvedilol or 50 mg twice daily of metoprolol tartrate. Doses were escalated at weekly intervals to achieve target BP control per recommended guidelines. Patients were followed-up for 35 weeks with no long-term follow-up planned. The primary end point was difference in HgbA1c after 5 months of maintenance BP therapy with either carvedilol or metoprolol tartrate. Secondary end points were BP control, insulin resistance, lipoprotein particle profiles, microalbumiuria, albumin/creatinine ratios, and withdrawals due to worsening glycemic control.

Results 

A total of 1235 patients were randomized to receive carvedilol (498 patients) or metoprolol tartrate (737 patients). Metoprolol increased baseline HgbA1c by 0.15% and carvedilol patients had no appreciable increase in HgbA1c (0.02%). More patients in the metoprolol arm withdrew due to worsening glycemic control (2.2%) than in the carvedilol arm (0.6%). HOMA-IR (a validated measure of glycemic control) was significantly improved in the carvedilol patients (P = .004). At the end of the study, carvedilol patients had more reduced albumin/creatinine ratios and better triglyceride panels than patients taking metoprolol. Hypertensive efficacy was similar in both groups, with more metoprolol patients experiencing bradycardia.

Intepretation 

Metoprolol and carvedilol are both effective therapies for hypertension in patients with diabetes. Patients taking carvedilol in this study had significantly better glycemic control, less insulin resistance, and less microabuminuria than patients taking metoprolol. There were no long-term data in GEMINI with regard to hard cardiovascular event end points; given the data that diabetics with better glycemic index have better cardiovascular outcomes, such studies must be done.

Study 

ISAR-SWEET (Randomized Clinical Trial of Abciximab in Diabetic Patients Undergoing Elective Percutaneous Coronary Interventions After Treatment With a High Loading Dose of Clopidogrel)

Presenter 

Dr. Julinda Mehilli, Deutsches Herzzentrum, Munich, Germany

Background 

Diabetic patients are at increased risk of complications after percutaneous coronary interventions (PCI). According to subset analyses, glycoprotein IIb/IIIa inhibitors have been demonstrated to improve cardiovascular outcomes in diabetic patients undergoing PCI.28., 29. However, no specifically designed randomized studies have addressed the value of abciximab, a glycoprotein IIb/IIIa inhibitor, in diabetic patients undergoing PCI, particularly after pretreatment with a high dose of clopidogrel. A recent trial found no additional clinical benefit associated with abciximab when low-risk, non-diabetic patients were pretreated with clopidogrel prior to PCI.30

The ISAR-SWEET trial was designed to evaluate whether abciximab had a favorable impact on the outcome of diabetic patients undergoing elective PCI after pretreatment with a 600-mg loading dose of clopidogrel.31 The investigators sought to compare a strategy of abciximab (standard dose 0.25 mg/kg bolus followed by 0.125 mcg/kg/min infusion) and unfractionated heparin with a strategy of placebo and heparin in 701 stable diabetic patients scheduled for elective PCI of a native vessel. Patients with high-risk markers of injury, recent myocardial infarction, angiographically visible thrombus, and left ventricular dysfunction were excluded. All patients received aspirin and a 600-mg loading dose of clopidogrel at least 2 hours before PCI. Clopidogrel therapy was continued for at least 6 months in all patients. Follow-up angiography was scheduled at 6 to 8 months following index PCI. The primary end point of the study was the cumulative incidence of death from any cause and myocardial infarction during the first 12 months after randomization. Secondary end points included the incidence of binary angiographic restenosis and the incidence of target lesion revascularization.

The trial was designed to assess the value of abciximab in this setting and a priori assumed a 50% relative risk reduction in the primary end point in the abciximab arm compared with the placebo arm. This was a double-blind, randomized, placebo-controlled trial and was supported by local institutional grants.

Results 

The 2 study arms had comparable baseline demographic, clinical, and angiographic characteristics. Of note, 29% of study patients had insulin-treated diabetes mellitus and 84% had multi-vessel coronary artery disease. The median time interval between clopidogrel administration and PCI was approximately 6 hours.

The primary end point of death and myocardial infarction at 1 year was 8.3% in the abciximab group and 8.6% in the placebo group (P = .91), with a relative risk of 0.97 (95% CI 0.58-1.62). There were no differences in the individual outcomes of the composite end point among the study groups. Follow-up angiography was obtained in 80% of eligible patients and the incidence of angiographic restenosis was 28.9% in the abciximab group and 37.8% in the placebo group (P = .01), with a relative risk of 0.76 (95% CI 0.62-0.94). The incidence of target lesion revascularization was 23.2% in the abciximab group and 30.4% in the placebo group (P = .03). In terms of adverse events, there appeared to be a trend toward a lower incidence of bleeding, thrombocytopenia, and transfusion in the placebo group, although these differences did not reach statistical significance.

Interestingly, subgroup analysis of patients treated with bare metal stents showed a reduction in the risk of angiographic restenosis in the abciximab group over the placebo group (31.9% vs 40.2%, P = .04). There was no difference in angiographic restenosis among the small proportion of patients treated with drug-eluting stents (10% of all PCI patients). Among insulin-treated patients, angiographic restenosis was 32.7% in the abciximab group and 41.5% in the placebo group (P = .19).

Interpretation 

In stable diabetic patients undergoing elective PCI after pretreatment with 600 mg of clopidogrel at least 2 hours before the procedure, a strategy of abciximab is not superior to placebo in the 1-year cumulative incidence of death and myocardial infarction. Abciximab did, however, reduce the risk of angiographic restenosis and target lesion revascularization in this diabetic population.

This study's important conclusions will need further validation from larger trials. Future trials also need to evaluate a similar strategy in diabetic patients with high-risk clinical and angiographic features and those treated with insulin regimens.

Study 

Effect of Rimonabant on Weight Reduction and Weight Maintenance: Rimonabant In Obesity-North America (RIO-NA) Trial

Presenter 

Dr. F. Xavier Pi-Sunyer, St. Luke's/Roosevelt Hospital Center, Columbia University, New York, NY

Background 

Epidemiologic studies have shown that increased body weight is associated with excess morbidity and mortality.32 Among adults, overweight and obesity substantially elevate the risk of illness from heart disease, metabolic syndrome, and diabetes. There has been an epidemic of obesity over the last 2 decades, and data from the 1999-2000 CDC Survey indicate that overweight and obesity have continued to increase in the United States, with estimates of 65% of adults being overweight and 31% obese.33 Obesity seems to be more prevalent among women, particularly in minority populations. Therefore, as obesity represents a growing major public health problem, strategies that provide weight reduction/maintenance and its benefits are needed.

The newly discovered endocannabinoid system is thought to play a key role in both central and peripheral regulation of energy balance, lipid metabolism, and insulin sensitivity, as well as in the control of tobacco dependence. Rimonabant is the first selective cannabinoid type 1 (CB1) blocker developed and established to be efficacious in managing cardiovascular risk factors such as obesity and smoking. It has been tested in a large program of phase III trials conducted in >6600 overweight/obese patients. Two recently completed trials including patients with untreated dyslipidemia (RIO-Lipids) or with or without comorbidities (RIO-Europe) have demonstrated that treatment with rimonabant 20 mg provided a significant weight reduction and improvements on both lipid and glycemic profiles relative to placebo, as well as a favorable safety profile.

The RIO-NA was a randomized, double-blind, placebo-controlled, parallel-group, fixed-dose study conducted on 3040 patients in Canada and the USA over a 2-year period to assess the effect of rimonabant on weight reduction and weight maintenance. Patients were eligible if they had a body mass index (BMI) ≥30 kg/m2 or a BMI ≥27 kg/m2 with comorbidities. Diabetic patients were not included. Overall, 80% were women, with mean age <50 years, BMI 37.6 kg/m2, weight 104.4 kg, and waist circumference 105.8 cm. They were prescribed a mildly hypocaloric diet and entered a 4-week placebo run-in period before being randomized to receive either rimonabant 5 mg, rimonabant 20 mg, or placebo for 1 year. After 1 year, patients in the rimonabant arms were re-randomized to receive either the same dose of rimonabant or placebo. The primary efficacy end points were weight loss, weight maintenance at 1year, and prevention of weight regain between the first and second years. The investigators also analyzed lipid concentrations and fasting glucose/insulin homeostasis as secondary end points.

Results 

Among completers at 2 years, 62.5% of patients treated with rimonabant 20 mg for 2 years lost >5% of initial body weight, compared with 36.7% of those on rimonabant 5 mg, and 33.2% of those on placebo (P < .001). In addition, 33% treated with rimonabant 20 mg lost >10% of initial weight versus 20% on rimonabant 5 mg and 16.4% on placebo (P < .001). At 2 years, the reduction in waist circumference was 8 cm, 4.9 cm, and 3.8 cm for those treated with rimonabant 20 mg, 5 mg, or placebo, respectively (P < .001), which represented around 1-cm reduction per 1 kg lost.

Regarding the metabolic abnormalities, HDL cholesterol levels increased by 24.5%, 15.6%, and 13.8% in the high-dose, low-dose rimonabant, and placebo groups, respectively (P < .001). Triglycerides were reduced by 9.9%, 5.9%, and 1.6% in patients treated with rimonabant 20 mg, 5 mg, and placebo, respectively (P < .05). In addition, the reduction in the proportion of patients with metabolic syndrome from baseline was greater at 2 years among those treated with rimonabant 20 mg relative to rimonabant 5 mg and placebo-treated patients (P < .001).

Interpretation 

The key findings are (1) treatment with rimonabant 20 mg for 2 years significantly reduced weight and abdominal fat and improved overall metabolic profile; (2) the treatment efficacy observed at 1 year was sustained at 2 years with a good safety profile; (3) these 2-year trial results are practically be superimposed with both the RIO-Lipids and RIO-Europe results, showing a strong consistency across the whole Rimonabant In Obesity program.

Therefore, treatment with rimonabant 20 mg for 2 years is a safe and efficacious strategy to help obese patients lose weight and abdominal fat as well as improve metabolic abnormalities, thus reducing the cardiovascular risk factors.

Study 

SHIELD (SHock Inhibition Evaluation with azimiLiDe): Placebo-Controlled, Randomized Clinical Trial of Azimilide for Prevention of Ventricular Tachyarrhythmias in Patients With an Implantable Cardioverter Defibrillator

Presenter 

Dr. Paul Dorian, St. Michael's Hospital, Toronto, Ontario, Canada.

Background 

Implanted cardioverter defibrillators (ICDs) have improved mortality and quality of life for patients with sustained ventricular tachycardia (VT) or ventricular fibrillation (VF).34., 35. Many patients with ICDs, however, continue to experience transient ventricular arrhythmias that cause disabling symptoms and induce painful ICD shocks. Therapeutic options for these patients are limited due to the high rates of adverse effects and many contraindications associated with currently available antiarrhythmic medications.

Azimilide dihydrochloride is an investigational antiarrhythmic drug that inhibits the rapid (IKr) and the slow (IKs) components of a repolarizing cardiac potassium current (Ik), thus prolonging the cardiac action potential and refractory periods. Previous studies have also shown that azimilide is effective for supraventricular arrhythmias,36 and a recent pilot trial of 172 patients demonstrated that azimilide reduces ICD shocks in patients who are at risk for recurrent VT/VF.37

In a randomized, double-blind, placebo-controlled study of 633 international patients with ICDs, this study assessed the safety and efficacy of azimilide in reducing recurrent tachyarrhythmia symptoms and ICD shocks. Patients were eligible for enrollment if they had a history of spontaneous sustained VT, or cardiac arrest/VF with an EF <40%; patients with a pre-existing ICD who experienced shocks due to spontaneous VT or VF were also included. Exclusion criteria included NYHA class IV heart failure, unstable angina, MI <30 days, prolonged QTc, or major illness. Participants were stratified by geographic regions and β-blocker use, and randomized in a 1:1:1 fashion to placebo, azimilide 75 mg, or azimilide 125 mg. The primary end points were all-cause shocks and all-cause shocks plus symptomatic tachyarrhythmias terminated by anti-tachycardia pacing (ATP); a secondary end point of all appropriate ICD therapies was also evaluated. The study was supported by Procter and Gamble Pharmaceuticals.

Results 

In an intention-to-treat analysis where no patients were lost to follow-up, the SHIELD investigators demonstrated that azimilide significantly reduced total all-cause shocks plus symptomatic VT terminated by ATP. With the 75-mg dose, the relative risk reduction was 57% (hazard ratio [HR] 0.43, 95% CI 0.26-0.69, P = .0006). With the 125-mg dose, the relative risk reduction was 47% (HR 0.53, 95% CI 0.34-0.83, P = .0053). The other primary end point of all-cause shocks did not reach significance with either dose of azimilide. For the secondary end point of all appropriate ICD therapies, both the 75-mg and 125-mg dose of azimilide achieved significant reductions (HR 0.52, 95% CI 0.30-0.89, P = .017; HR 0.38, 95% CI 0.22-0.65, P = .0004, respectively). Additionally, 75 mg of azimilde significantly reduced cardiac-related hospitalization and emergency room (ER) visits. In regard to adverse events, 5 patients receiving azimilide and 1 placebo patient experienced torsades de pointes. One patient receiving azimilide had severe but reversible neutropenia.

Interpretation 

Results from the SHIELD study demonstrate that azimilide is a safe and effective drug to reduce the recurrence of VT/VF terminated by shocks or ATP in patients with ICDs. Importantly, the 75-mg dose of azimilide also reduced cardiac-related hospitalizations and ER visits.

Although very promising, the results of this study should be considered in the context of its highly selected patient population and the fact that it was a relatively short study with a follow-up of only 1 year.

Study 

ARBITER-2: ARterial Biology for the Investigation of the Treatment Effects of Reducing Cholesterol: A Double-Blind, Placebo-Controlled Study of Long-Acting Niacin on Atherosclerosis Progression in Secondary Prevention Patients Treated with Statins

Presenter 

Dr. Allen J. Taylor, Walter Reed Army Medical Center, Washington DC

Background 

It is well established that pharmacologic lowering of low-density lipoprotein cholesterol (LDL-C) reduces cardiovascular events. Low high-density lipoprotein cholesterol (HDL-C) is a known cardiovascular risk factor, but there is uncertainty about the incremental benefit of increasing HDL-C in preventing cardiovascular events. Niacin has been shown to improve coronary heart disease (CHD) outcomes,38 but no trial has yet demonstrated the superiority of combination therapy with a statin and niacin over statin monotherapy. To address these issues, the ARterial Biology for the Investigation of the Treatment Effects of Reducing Cholesterol (ARBITER-2) study investigators performed a randomized, double-blind, placebo-controlled study of long-acting niacin in addition to background statin therapy for secondary prevention in patients with known CHD. Carotid intima media thickness (CIMT) was used as an intermediate outcome measure to address the primary objective of comparing the effect of extended release (ER) niacin versus placebo on change in CIMT in patients with known CHD already on statin therapy.

The ARBITER-2 study enrolled patients from December 2001 to May 2003. Inclusion criteria were age >30 years; stable coronary artery disease (CAD); statin therapy with LDL-C <130 mg/dL; and HDL-C <45 mg/dL. The study design consisted of a 1:1 randomization to niacin 500 mg nightly for 1 month, followed by 1000 mg nightly for another 11 months, versus placebo. Baseline labs and CIMT were done at enrollment and again at 12 months. At 3- and 6-month follow-up, labs and adherence to study medications were assessed. CIMT was performed using carotid ultrasound by 1 blinded sonographer using a single 8-mHz probe, measured at the far wall of the distal common carotid, and was a mean value of bilateral measurements. The primary study end point was change in mean CIMT. Secondary end points included efficacy and safety end points, and composite clinical cardiovascular events.

Results 

Results of the ARBITER-2 study were published online November 10, 2004.39 The ARBITER-2 study randomized 167 patients to ER niacin versus placebo. Baseline characteristics were similar between the 2 groups, with a male predominance, mean age of 68 years, and mean CIMT of approximately 0.9 mm. Baseline therapies were also similar between groups; 100% of patients were treated with statins by study design, with 93% of patients treated with simvastatin. The mean duration of statin therapy was 4.5 years, with an average dose of 35 mg daily. There was a high prevalence of other CAD pharmacotherapies, including β-blockers (79%) and aspirin (86%). Study participants were encouraged to discontinue anti-oxidant vitamin supplements. Mean lipid values were likewise similar between the 2 groups, with a mean LDL-C of 91 mg/dL (placebo) and 87 mg/dL (treatment group). Mean HDL-C was approximately 40 mg/dL in both groups. After 12 months of therapy, LDL-C was unchanged in both groups. HDL-C had increased from 39 to 47 mg/dL in the ER niacin-treated group (21% increase, P = .002), and remained unchanged in the placebo arm. Serum triglycerides decreased in the ER niacin group and remained unchanged in the placebo arm.

CIMT progression decreased 68% in patients treated with ER niacin compared with placebo (0.014 mm ± 0.011 mm, P = .23 for CIMT progression in the niacin group, versus 0.044 mm ± 0.011 mm, P = .001 for CIMT progression in the placebo group; P = .08 for comparison between groups). Of the placebo group, 25.4% showed stabilization or regression of CIMT, compared with 38.5% in the ER niacin-treated group (P = .09). A non-prespecified, post-hoc analysis stratified by the presence of diabetes mellitus or metabolic syndrome found that the most significant decrease in CIMT progression was in the subgroup without diabetes or metabolic syndrome. For the composite cardiovascular events secondary end point there was a statistically insignificant 60% decrease in events in the ER niacin group compared with placebo (9.6% vs 3.8% for the placebo and treatment groups, respectively; P = .20). Secondary efficacy and safety end points revealed tolerance and lack of major adverse side effects in both group, most notable for no occurrence of transaminitis or myositis in either group. Drug adherence was high (>90% in both groups).

Interpretation 

ARBITER-2 is the first clinical trial to evaluate the incremental benefit of combination therapy with niacin and a statin over statin monotherapy for secondary prevention in patients with CAD. The results reveal that, in a population of patients with CAD already on statin therapy, addition of ER niacin slowed progression of CIMT by 68%. Furthermore, placebo-treated patients with a mean HDL-C of 40 mg/dL showed a progression of CIMT despite a mean LDL-C of 86 mg/dL. A statistically insignificant trend towards fewer cardiovascular events in patients treated with combination therapy was also noted. ER niacin was found to be safe and well tolerated, with a high adherence rate. ARBITER-2 is a well designed, well conducted study, suggesting that focus on LDL lowering alone is insufficient in the management of lipids for secondary prevention of CAD, and supporting the notion that lowering LDL-C to <100 mg/dL is insufficient. The issue is raised, however, of whether niacin in addition to statin therapy would also be beneficial in patients treated aggressively to an LDL-C <70 mg/dL, as currently offered as a therapeutic option in the most recent National Cholesterol Education Program (NCEP) report.40 Although the ARBITER-2 results must be viewed in light of its evaluation only of intermediate end points, they are compelling for the consideration of the addition of niacin in lipid management for secondary prevention of CAD in patients with moderately low HDL-C levels. A larger clinical trial with a primary end point of cardiovascular events is justified.