Emergency Medicine

Emerging Issues in Cardiology

The author extracts key points from some of the more recent clinical trials that have implications for the care of patients with acute coronary syndromes, unstable angina, myocardial infarction including non-STEMI, or coronary heart disease. The use of angiotensin receptor blockers is also discussed.

By Elliot Rapaport, MD

Practice patterns in cardiology today are highly reliant on evidenced-based medicine derived from the results of numerous randomized clinical trials, as well as overviews or meta-analyses when trials are small or the results conflicting. This article is an update summarizing results from some of the more recent clinical trials.

ACUTE CORONARY SYNDROMES

Approximately 8 million people present in the emergency departments of hospitals in the United States each year complaining of chest pain. In 5 million of these patients, their description of the chest pain suggests an ischemic origin, which results in roughly 2.3 million hospitalizations for an acute coronary syndrome (ACS). Although an immediate history and physical examination are the routine first steps in the evaluation of these patients, prompt triage demands that a 12-lead ECG should be obtained without delay. This allows immediate identification of those who are undergoing an ST-segment elevation acute myocardial infarction (STEMI). If the ST-segment elevation and chest pain do not resolve immediately after the patient has chewed a 325-mg aspirin tablet and a sublingual nitroglycerin tablet has been administered, the patient should be promptly transported to the catheterization laboratory for angioplasty, provided an experienced interventionalist and team are readily available. While the patient is awaiting transport, a beta blocker should be administered intravenously and either enoxaparin or unfractionated heparin should be started.

Because primary angioplasty results in a better short-term outlook than thrombolysis, it is the preferred approach. A recent meta-analysis of 23 trials comparing these two approaches showed a distinct survival benefit and reduction in adverse cardiac events with angioplasty. If no catheterization facility is available at the hospital and there are no plans to transfer the patient to a facility where percutaneous cardiac intervention (PCI) can be performed, or if access to an existing catheterization lab will be substantially delayed, thrombolytic therapy should be started immediately. The DANAMI-2 study in Denmark of 1,572 STEMI patients demonstrated that the combined endpoint of death, reinfarction, or stroke at 30 days was significantly less in patients promptly transported to a referral center (96% arrived within two hours) for primary angioplasty, compared to that observed in patients randomized initially to thrombolysis with an accelerated alteplase infusion (8.5% versus 14.2%, respectively).

Another alternative, if a significant delay is anticipated, is to start thrombolytic therapy while arranging for transport to another facility where angioplasty can be carried out. The results from such an approach appear to be comparable to those with immediate primary angioplasty. The infusions of alteplase or streptokinase used in past years to accomplish thrombolysis have given way to third-generation thrombolytics that permit bolus injections. Generally, most patients receive either two boluses of reteplase 30 minutes apart or, more commonly today, a single weight-optimized bolus of tenecteplase. Based on the recent ASSENT III trial and similar results from the ENTIRE-TIMI 23 trial, my institution has adopted the strategy of combining full-dose tenectaplase with enoxaparin, which appears to offer better results compared to the combined use of tenecteplase and unfractionated heparin.

If a patient fails to respond clinically, as evidenced by persistent, significant chest pain or failure of the ST segments to return promptly to baseline, he or she should be taken to the catheterization laboratory as soon as possible. Similarly, if either significant chest pain or ST-segment elevation recurs, immediate catheterization is generally indicated.

Combination therapy using half-dose tenecteplase with abciximab in ASSENT III and half-dose reteplase with abciximab in GUSTO V produced comparable benefits to those just described. However, these therapies are associated with an increase in major bleeding complications and have not, therefore, been widely accepted. They are particularly likely to prove harmful in the very elderly patient.

UNSTABLE ANGINA AND NON-STEMI

Over 1.4 million patients hospitalized with ischemic chest pain each year experience unstable angina or an evolving non-STEMI. A great deal of variability exists in the way such patients are managed from hospital to hospital, although the trend is toward pursuing an early invasive strategy. The efficacy of such a strategy has been demonstrated in three randomized clinical trials in recent years: the FRISC-2 and TACTICS TIMI-18 trials, and, most recently, the RITA-3 trial.

RITA-3 re-explored the question of whether an invasive strategy or a conservative approach is better in managing patients with unstable angina or a non-STEMI. This study from the United Kingdom involved the randomization of 1,810 patients considered to be at moderate risk for death. The two primary endpoints were the composite rate of death, nonfatal MI, and refractory angina at four months and death or nonfatal MI at one year. Of note was the fact that patients in both groups were given enoxaparin, aspirin, and beta blockers, if tolerated. At four months, the primary endpoint had occurred in 9.6% of the intervention cohort and 14.5% in the patients randomized to a conservative strategy. This benefit was essentially a reflection of a marked decrease in the rate of refractory angina in the intervention group. At one year, the rate of death or MI was similar in both groups.

These three trials suggest a better outcome at six months after an invasive strategy, compared to a conservative strategy that waits until recurrent or continuing ischemia occurs despite medical management before the patient is sent to the catheterization lab. In addition, there have been several developments that encourage direct transport of a patient soon after admission to the catheterization lab or even directly from the emergency department. These include improvements in PCI procedures (such as bare metal stents and, more recently, drug-eluding stents), use of clopidogrel (not only as a loading dose prior to PCI but also for up to one year after stenting, as in the CREDO trial), and use of bivalirudin instead of heparin and a glycoprotein IIb/IIIa inhibitor during PCI, as in the REPLACE-2 trial. (The glycoprotein IIb/IIIa inhibitor would still be available as a "bail-out," if needed.)

Nevertheless, it should be pointed out that the long-term benefit is almost exclusively in the prevention of recurrent ischemia or MI, or both. The graph below shows the results from RITA-3 added to an earlier overview published by Barnett and colleagues. When you look at the reported six-month to one-year mortality in the five major randomized trials examining a conservative compared with an invasive management strategy in 8,822 patients, there are actually a few more deaths in the invasive arm.

Less is more? Pooled analysis of the TIMI-3B, VANQWISH, FRISC-II, TACTICS-TIMI 18, and RITA-3 trials, representing a total of 8,822 patients, finds a slightly greater number of deaths among those treated invasively.

A number of emergency departments have developed protocols for the administration of glycoprotein IIb/IIIa inhibitor infusions in patients presenting with an ACS and awaiting definitive disposition. Although such infusions are appropriate for patients who will be transported to the catheterization lab, they are not indicated routinely in the patient for whom medical management is initially contemplated. In the large GUSTO IV trial, where patients were more or less universally treated medically, both 24- and 48-hour infusions of abciximab compared to placebo were clearly detrimental in terms of 30-day outcome.

For this reason, the American College of Cardiology/ American Heart Association guidelines for ACS management have assigned a class 3 designation (or contraindication) for use of abciximab under these circumstances. This is somewhat puzzling in light of earlier evidence that showed that the best results in patients undergoing a PCI were in those trials where abciximab was used compared to placebo in the catheterization lab.

Other trials of a glycoprotein IIb/IIIa inhibitor, particularly tirofiban and eptifibitide, in which the patient had a PCI performed, also demonstrated an advantage with adjunctive use of the inhibitor. However, this is not the case in those trials where initial medical management was mandated. In those six trials (PRISM, PRISM-PLUS, PARAGON A, PARAGON B, PURSUIT, and GUSTO IV), where a prolonged infusion of a glycoprotein IIb/IIIa inhibitor was compared to placebo, a meta-analysis by Boersma and colleagues showed only a minor overall benefit in 30-day death or nonfatal MI outcome.

Furthermore, a more recent meta-analysis of these trials by Roffi and colleagues that looked only at 30-day mortality revealed no benefit whatsoever in the overall ACS population receiving a glycoprotein IIb/IIIa infusion. Interestingly, however, a reduction in 30-day mortality was seen in the 6,458 enrollees who were diabetic (4.6% versus 6.2%), but there was no mortality benefit among the 23,072 enrolled nondiabetics (3% versus 3%).

The results of these trials suggest that glycoprotein IIb/IIIa infusions should be reserved for patients on their way to or already in the catheterization lab. Otherwise, their use should probably be restricted to diabetic patients with recurrent chest pain being managed medically who cannot be promptly transported to the catheterization lab. It should also be kept in mind that the clinical trials of the glycoprotein IIb/IIIa inhibitors were carried out before the CURE and PCI-CURE trial results were known. As a result of these trials, we now have an alternative approach in antiplatelet medical management with the use of daily oral clopidogrel and aspirin for up to nine months, which has been shown to be effective in reducing adverse cardiac events in the ACS patient.

Furthermore, the CURE investigators were able to demonstrate a significant benefit as early as 24 hours after clopidogrel administration in the rate of cardiovascular death, MI, stroke, and refractory angina. This finding was reinforced by the benefit found, in terms of 28-day outcome, in the more recent CREDO trial with administration of a loading dose of clopidogrel between 6 and 24 hours before PCI, compared with the failure to demonstrate a benefit when it was given within 6 hours prior to the PCI. Recent further analysis of the CURE data has revealed that a benefit was seen in low-risk as well as intermediate- and high-risk patients, based on their TIMI scores, with the use of clopidogrel. This suggests that early administration of clopidogrel should be part of routine management.

CORONARY HEART DISEASE

Several years ago, the HOPE study established the benefit of the ACE inhibitor ramipril in preventing death, MI, and stroke in high-risk patients with presumed normal left ventricular (LV) systolic pump function who were over the age of 55. The original inclusionary criteria required either the presence of atherosclerosis in one or more vascular beds or the presence of diabetes and one other risk factor.

Recently, the EUROPA trial results complemented the HOPE findings. EUROPA was a trial of 12,218 patients with stable coronary heart disease (CHD) who were at least 18 years of age and without clinical evidence of heart failure. They were randomized to receive either 8 mg/d of the ACE inhibitor perindopril or matching placebo and were followed for a mean 4.2 years. The study found that 10% of the placebo patients and 8% of the perindopril cohort experienced the primary endpoint of cardiovascular death, MI, or cardiac arrest. This is equivalent to a 20% relative risk reduction. These benefits were observed across a broad spectrum of patients, including all age groups and in patients with and without hypertension, diabetes mellitus, or previous MI, and despite significant therapy with antiplatelet drugs, lipid-lowering agents, and beta blockers.

In my judgment, these findings reinforce the concept that all patients with atherosclerotic vascular disease and all adults with diabetes should be on an ACE inhibitor, preferably ramipril or perindopril. (As of this writing, perindopril is awaiting FDA approval for this use, although it is an approved antihypertensive drug.)

The MRC/BHF Heart Protection Study (HPS) was a landmark study involving 20,536 adults in the United Kingdom, aged 40 to 80, who were high-risk by virtue of having atherosclerotic vascular disease, diabetes, or treated hypertension. They were randomized to receive either 40 mg/d of simvastatin or placebo. After five years, all-cause mortality was reduced from 14.7% to 12.9%, CHD mortality from 6.9% to 5.7%, nonfatal or fatal stroke from 5.7% to 4.3%, and fatal or nonfatal MI from 11.8% to 8.7%. Again, this benefit was seen across a wide spectrum of participants, including those with diabetes, peripheral vascular disease, CHD, cerebrovascular disease without diagnosed CHD, women as well as men, and those over and under 70 years of age at baseline.

Of particular importance was the observation that the relative risk reduction with simvastatin was similar regardless of whether a participant's low-density lipoprotein (LDL) cholesterol at baseline was under 100 mg/dl, between 100 and 129 mg/dl, or 130 mg/dl or higher. These results clearly indicate that there will have to be a prompt revision of the current U.S. ATP III guidelines relating to both the LDL cholesterol level at which statin therapy should be initiated and the LDL cholesterol goal to be achieved. In the absence of an obvious contraindication, it seems that all adults with diabetes as well as patients of any age with atherosclerotic vascular disease should be on chronic statin therapy regardless of their LDL cholesterol level. I suspect that the target level to be achieved will be closer to 60 or 70 mg/dl, as opposed to the current goal of less than 100 mg/dl. Several randomized trials are currently addressing this issue.

Additional support for starting statin therapy in the high-risk patient even in the absence of known atherosclerotic vascular disease was provided by the ASCOT-LLA trial. This was an arm of the overall ASCOT trial that is examining two different strategies: a primary beta blocker versus a primary calcium channel blocker for managing high-risk hypertensive patients. Inclusionary criteria in this overall trial involving 19,432 patients required at least three or more risk factors. Of these patients, 10,305 with a total cholesterol level of less than 250 mg/dl were randomized to receive either 10 mg of atorvastatin or placebo daily. Atorvastatin produced a 23% reduction in total cholesterol and a 32% reduction in LDL cholesterol compared to baseline. Although the overall trial is continuing, the lipid arm of the trial was stopped prematurely because of atorvastatin's efficacy. Blood pressure was reduced identically by 26/15 mm Hg in both groups; however, there was a 36% risk reduction in the primary endpoint of nonfatal MI and CHD mortality with the use of atorvastatin. Additionally, total coronary events were reduced by 29% and fatal and nonfatal episodes of stroke by 27%.

POST-MI PATIENTS

The MADIT II trial looked at whether patients who were more than 30 days post-MI but with marked impaired LV contractility (ejection fraction equal to or less than 30%) and no ventricular arrhythmia would benefit from implantation of an automatic implantable cardioverter defibrillator (AICD). Entry criteria did not require a prior electrophysiologic study. The trial compared patients who had an AICD implanted and also were receiving conventional treatment to conventional treatment alone. The trial was stopped prematurely because of a dramatic 30% risk reduction in mortality among those randomized to receive the AICD.

A post-study analysis revealed that the greatest benefit was seen in those patients who had a QRS interval of more than 0.12 seconds at baseline. Medicare has seized on this to authorize reimbursement only for those having this additional finding and not just the inclusionary criteria of the trial. In some ways, this is understandable, in light of the fact that roughly 250,000 patients will qualify for implantation of an AICD each year in the United States. Without the QRS interval stipulation, the costs of these procedures would run to $6 to $7 billion a year, based on current expenditures.

It should also be pointed out that the need for hospitalization for heart failure was higher in the AICD cohort than among the usual-care patients. This probably reflects the fact that an arrhythmia-induced death is common in heart failure patients. Presumably, the AICD prevented such a death, making it more likely that they would be hospitalized as their heart failure progressed.

The CIBIS II, MERIT-HF, and COPERNICUS trials several years ago clearly established the benefit that can be achieved with the use of a beta blocker in class II-IV heart failure. Where evidence has been lacking is in the class I patient—that is, the patient with a recent MI who has residual impaired LV contractility. The CAPRICORN trial enrolled 1,959 patients who had a recent MI and an ejection fraction of 40% or less. These patients were randomized to receive either a progressively increasing dose of carvedilol (up to a target dose of 25 mg twice daily) or placebo. The results showed a significant relative risk reduction of 23% in all-cause mortality. Other endpoints such as cardiovascular mortality and nonfatal MI were also reduced. This effectively closes the loop, indicating that beta blocker therapy has a benefit in all patients with heart failure ranging from the class I asymptomatic patient with LV dysfunction to the class IV patient with heart failure exhibiting signs and symptoms at rest.

The MAGIC trial reexamined the issue of intravenous (IV) magnesium therapy for the high-risk patient experiencing an acute STEMI. This trial was prompted by the seemingly contradictory findings in two large trials—LIMIT II, in which dramatic benefit was seen, and ISIS IV, in which no benefit was seen in a trial involving more than 50,000 MI patients. The MAGIC trial randomized 6,213 patients to receive either an IV bolus of 2 gm of magnesium sulfate over 15 minutes, followed by a 17-gm infusion over the next 24 hours, or a matching placebo. The primary endpoint was 30-day, all-cause mortality; the odds ratio for this endpoint turned out to be 1.0.

The results were not only negative for mortality but also for a variety of other endpoints involving pre-specified groups of patients, including those undergoing reperfusion and those who were not, those over age 65 eligible for reperfusion, and those in any age group ineligible for reperfusion. It is clear from the MAGIC trial that there is no justification for the routine use of IV magnesium in patients experiencing a STEMI. Its use should be restricted to those patients who have a specific indication for magnesium therapy.

The EPHESUS trial examined the effect of eplerenone, a selective mineralocorticoid aldosterone receptor blocker previously approved for use in hypertension, in patients who had experienced an MI 3 to 14 days earlier, were in heart failure with an LV ejection fraction of 40% or lower, and were mostly class II patients. The earlier RALES trial demonstrated improved survival in class III and IV heart failure patients who received the relatively nonselective aldosterone inhibitor aldactone. In EPHESUS, 6,632 patients were randomized to receive either eplerenone (with an eventual dose of 50 mg/d) or placebo for a mean follow-up of 16 months. Eplerenone reduced the rate of cardiovascular mortality or hospitalization for cardiac events, the primary endpoint, with an odds ratio of 0.87.

The benefit with eplerenone was observed even when patients were being well managed with beta blockers, ACE inhibitors, statins, or aspirin and had undergone reperfusion therapy. This agent has just been approved by the FDA for use in patients meeting the inclusionary criteria enumerated above. It should be noted that patients with a serum creatinine above 2.5 mg/dl or a potassium level over 5 mEq/dl were excluded from the trial.

ANGIOTENSIN RECEPTOR BLOCKERS

There has been considerable interest in defining the role of angiotensin receptor blockers (ARBs) in treating heart disease. Are they better or worse than an ACE inhibitor in settings where an ACE inhibitor has been demonstrated to be beneficial? Is the combination of an ARB and an ACE inhibitor superior to an ACE inhibitor alone? Recent clinical trials have addressed these questions as they relate to patients who have had an MI and those who are in heart failure. Angiotensin receptor blockers have also been compared recently to beta blockers in treating hypertension.

The OPTIMAAL trial compared the ACE inhibitor captopril to the ARB losartan in high-risk post-MI patients. A total of 5,477 patients who had experienced heart failure with their acute MI or displayed a new Q-wave anterior MI were randomized to receive either losartan 50 mg/d or captopril 50 mg three times daily. All-cause mortality was the primary endpoint. Over a mean follow-up of 2.7 years, the relative risk with losartin tended to be worse (1.13). Thus, there would appear to be no reason to substitute losartan for captopril in the post-MI patient unless the patient cannot tolerate the ACE inhibitor. It should be noted, however, that the 50 mg/d dose of losartan used in this trial is on the low side. Most cardiologists will prescribe 100 mg/d.

The LIFE trial compared losartan to the beta blocker atenolol in patients with hypertension who also displayed electrocardiographic evidence of LV hypertrophy. A total of 9,193 patients were randomized to one or the other treatment for at least four years. The primary endpoint was a composite of death, MI, or stroke. At the conclusion of the trial, the reduction in blood pressure was essentially the same in both treatment groups—30.2/16.6 with losartan and 29.1/16.8 with atenolol. However, losartan proved superior in reducing the primary endpoint (relative risk, 0.87). Interestingly, there was no significant difference in the rate of death or nonfatal MI. Instead, all of the benefit was related to a decrease in the rate of fatal and nonfatal stroke with losartan (relative risk, 0.75). The rate of development of new-onset diabetes was also lower in the losartan cohort.

The Val-HeFT trial examined the addition of the ARB valsartan compared to placebo in 5,010 patients receiving standard treatment for class II-IV heart failure. The two primary endpoints were mortality and mortality or morbidity (mainly hospitalization for heart failure). Mortality was similar in the two groups. However, the combined mortality or morbidity endpoint was lower in the valsartan group, reflecting its benefit in preventing the need for rehospitalization (relative risk, 0.87). Hospitalization for heart failure was significantly reduced in the valsartan arm (13.8% versus 18%). An unsettling finding, however, came out of a post-hoc subgroup analysis that looked at the addition of valsartan in patients who were already on both an ACE inhibitor and a beta blocker. In these patients, the placebo group had significantly less mortality than the valsartan cohort. This finding leaves clinicians reluctant to add valsartan to the treatment regimen of patients in heart failure unless they are not on a beta blocker or an ACE inhibitor.

Fortunately, another trial, CHARM, which tested the ARB candasartan in heart failure patients, reported its findings recently. This trial was actually a composite of three separate trials: candasartan or placebo in patients with preserved LV function (the CHARM-Preserved trial); candasartan or placebo in patients with impaired LV contractility who could not tolerate an ACE inhibitor (the CHARM-Alternative trial); and candasartan or placebo added to the regimens of patients with impaired LV contractility, all of whom were already on an ACE inhibitor (the CHARM-Added trial).

The CHARM-Preserved trial enrolled 3,023 patients with class II-IV heart failure who had an LV ejection fraction greater than 40%. The primary endpoint was cardiovascular death or hospital admission for heart failure. At a median three years, the unadjusted hazard ratio for this outcome was 0.89. Although cardiovascular death did not differ between the two cohorts, fewer patients were initially hospitalized among the candasartan-treated patients. It would seem reasonable, therefore, to consider using candasartan when managing patients with so-called diastolic heart failure.

The CHARM-Alternative trial enrolled 2,028 patients with heart failure and an LV ejection fraction of less than 40% who were being managed without an ACE inhibitor because they could not tolerate the drug. Most of these cases of intolerance were due to intractable coughing. The primary endpoint again was a composite of cardiovascular death or hospitalization for heart failure. After a median follow-up of 33.7 months, both of these outcomes were significantly less frequent in the candasartan cohort, and the unadjusted hazard ratio for the primary endpoint was 0.77. These results indicate that there is a benefit to treating patients with an ARB when they have heart failure associated with impaired LV contractility but cannot tolerate an ACE inhibitor.

The CHARM-Added trial enrolled 2,548 patients with class II-IV heart failure and an LV ejection fraction of less than 40%, all of whom were taking an ACE inhibitor. Patients were randomized to receive either candasartan or placebo in addition to the ACE inhibitor and were followed for a median 41 months. The primary endpoint again was the composite of cardiovascular death or hospitalization for heart failure, which was experienced by 42% of the placebo group and 38% of the candasartan group. The unadjusted hazard ratio was 0.85. Thus, this trial, unlike the Val-HeFT trial, demonstrated a benefit in adding an ARB to the regimen of a patient already receiving an ACE inhibitor. Moreover, the results were not significantly different in the 55% of patients who were also on a beta blocker, in addition to the ARB and ACE inhibitor, compared to those who were not.

INFLUENTIAL TRIALS

This compendium of recent trials in cardiology has focused on those that are likely to immediately influence clinical practice. Failure to include others is not meant to convey any implication regarding their importance or eventual impact on practice. Rather, it is a simple recognition that it would take volumes to discuss them all in a meaningful way.