Improving Acute Coronary Syndrome Care: The ACC/AHA Guidelines and Critical Pathways

Approximately 1.42 million patients are admitted every year to hospitals in the United States with unstable angina and non-ST-segment elevation myocardial infarction (UA/NSTEMI). Over the past several years, advances in the evaluation and management of this large patient population have brought forth numerous effective medical therapies such as antiplatelet agents, cholesterol-lowering drugs, beta-blockers, and other agents. To help improve the treatment of UA/NSTEMI, the American College of Cardiology (ACC) and the American Heart Association (AHA) developed guidelines for the diagnosis and management of these patients which were recently updated. Just eighteen months after this revision, the guidelines were updated again following the publication of several landmark trial results.¹ The guidelines were established by a panel of experts who make recommendations on appropriate use of medications and interventions based on evidence from randomized clinical trials. ACC/AHA Guidelines. The recommendation to “target” the number of antithrombotic agents to the risk of the patient is illustrated in Figure 1. Of note is the fact that the addition of a glycoprotein IIb/IIIa inhibitor on top of aspirin, heparin, and clopidogrel, has actually been downgraded from a Class I to a Class IIa classification due to the absence of direct randomized trial data on this combination of therapies.¹ The primary question centers on the upstream use of glycoprotein IIb/IIIa inhibition, which, in the view of many, should be based on the patient’s risk factors. Figure 2 illustrates the huge benefit (approximately 70%) conferred by early IIb/IIIa inhibitor use in the CAPTURE, PRISM, PARAGON B, and PRISM-PLUS trials, whereas no benefits were observed in the patients with negative troponin values.^2–5Clopidogrel. The story is different for clopidogrel. The application of the TIMI Risk Score in the CURE trial actually demonstrated for the first time that the low-risk patient population also derived benefit from clopidogrel use.⁶ Twelve-month outcomes showed that there was a statistically significant reduction in the rate of cardiovascular death, myocardial infarction, or stroke in all three risk groups, including the low-risk patient group. Given the pharmacologic difference between the IIb/IIIa inhibitors and clopidogrel — and aspirin for that matter, which has been shown to benefit all patient groups in the antiplatelet trials — it may be revealed that aspirin and clopidogrel decrease platelet activation, which would decrease the number of activated platelets that are circulating in the bloodstream. However, since IIb/IIIa receptor blockers inhibit more directly into the thrombus, they would have a more direct antithrombotic effect on the thrombus, but would not decrease the overall propensity for thrombosis systemically. Placing these two therapies side-by-side reveals that IIb/IIIa inhibition is a great acute therapy in high-risk patients, while platelet activation inhibitors (aspirin and clopidogrel) are beneficial in both short- and long-term treatment for all patients. Balancing the timing of early treatment. In the CURE trial, approximately 75 events were prevented when the patient was treated with clopidogrel in the first two days, and would not have been prevented if clopidogrel treatment had been postponed (e.g., until cardiac catheterization). Much of the noise in the cardiology community against early treatment stems from the recently published observational study by Hongo et al.⁷ Hongo and colleagues looked at those patients who were given clopidogrel versus those who did not receive the drug pre-CABG. Those who were treated with clopidogrel were probably the highest risk patients, as opposed to the routine low-risk patients. The investigators observed higher rates of post-CABG bleeding in these high-risk patients based on various measures. The most frequently quoted statistic from this study is the “ten-fold increase in re-operation for bleeding.” On a cautionary note, however, we have been burned by observational data, like those from the above-cited article, several times in the recent past, notably in the areas of hormone replacement therapy,^8,9 vitamin E,¹⁰ or both.¹¹ Prior observational studies showed apparent benefits with differences in rates of endpoints between these groups, but these benefits did not appear in the randomized studies.^8–11 Data was examined from the TACTICS-TIMI 18 trial in which patients were not to receive upfront clopidogrel (the use of ticlopidine and clopidogrel by protocol was only post-stent).¹² Only a handful of patients (approximately 10–15 out of the 2,220 patients) in the TACTICS-TIMI 18 study received pre-treatment with ticlopidine or clopidogrel. Data was then examined on the patients in TACTICS-TIMI 18 who underwent CABG within five days of admission versus those who had the surgery after five days (Figure 3). The rate of major bleeding was four-fold higher in the group who underwent CABG early — and this was with no clopidogrel. However, had clopidogrel been given in this study, many would have attributed the excess bleeding to the fact that the clopidogrel didn’t wear off. The fact is, however, that the differences in the patients’ characteristics are what account for a large part of the differences in bleeding rates between the two groups. Thus, one must remember that not one, but many, confounding factors are involved in any “observation” of differences between two groups. Given that, it is necessary to return to the randomized data from the CURE trial to more accurately assess the effect of clopidogrel on bleeding in the setting of CABG (Figure 4).^13,14 In the CURE trial, no difference in TIMI major bleeding rates were observed as a whole, or among the patients who underwent CABG within five days of receiving the study drug.^13,14 There was an excess of bleeding, but this would fall into the TIMI minor bleeding category. Thus, the randomized data show a modest increase in TIMI minor bleeding rates, and little difference in TIMI major bleeding rates, which contrasts with the observational data that suggested a ten-fold increase in the rates of re-operation due to bleeding. The strength of the evidence from randomized data are what should drive treatment decisions as efficacy and safety issues are balanced. This is certainly under debate at Brigham and Women’s hospital, and it is fortunate that all the data are now available to assess this issue more accurately. One of the cases Jeff Popma will present here this evening will focus on the question of whether it is preferable to initiate clopidogrel treatment early or wait until post-procedure and perform an angiogram. Invasive versus conservative strategies. Figure 5¹⁵ shows results from six of the eight trials comparing invasive versus conservative strategies in the treatment of UA/NSTEMI. The TIMI 3 trial showed equal outcomes in terms of death or myocardial infarction and actually showed fewer re-hospitalizations.¹⁶ VANQWISH tilted the strategy somewhat away from the catheterization laboratory for a brief period, when mortality rates appeared higher with the invasive strategy.¹⁷ The MATE trial, which was conducted two years later, showed similar outcomes to those of TIMI 3.¹⁸ Beginning with FRISC II, the pendulum has swung back in favor of the invasive strategy.¹⁹ The boxes in Figure 6 are representative of the size of the trials in terms of the weight of their evidence. Jeffrey Popma and Spencer King both sat on the steering committee for TACTICS-TIMI 18, a trial that added to the weight of the evidence supporting an early invasive strategy.¹² These two concurrent, contemporary trials showed benefit conferred by early invasive treatment. Two small pilot trials in Europe^20,21 and the recent RITA-3 trial²² also showed dramatic benefits with the early invasive strategy. In total, 6,618 patients from five trials were shown to benefit from the invasive strategy; 1,674 patients demonstrated equal outcomes; and just 900 patients (VANQWISH trial) were shown to benefit from the conservative strategy (Figure 5). The invasive strategy thus has clearly proven beneficial. It has become apparent that catheterizing absolutely every patient is not necessary. As was observed with IIb/IIIa inhibition, the troponin and ST-segment deviation subgroups showed benefit with the invasive strategy in the higher risk patients and no apparent benefit in the lower risk group (Figure 7).^12,13 The TIMI risk score demonstrates the same results, with benefit observed in the 75% of patients who fall into the intermediate- and high-risk groups.^12,13 In terms of the ACC/AHA guidelines, there are now five baseline therapies for all patients: aspirin, clopidogrel, heparin or low-molecular weight heparin (with enoxaparin as the preferred antithrombin), beta-blockers, and nitrates (Figure 8).²³ Furthermore, invasive treatment and glycoprotein IIb/IIIa inhibitors are best targeted with risk stratification, with the benefit conferred to the higher risk patients. Long-term management post-ACS. For long-term management of acute coronary syndrome patients, the ACC/AHA guidelines now place much greater emphasis on starting the secondary prevention medications and on risk factor modification, which is the focus for this meeting. The ACC/AHA Guidelines recommend the following five classes of drugs: the combination of aspirin and clopidogrel, beta-blockers; lipid-lowering agents (recommended for all patients with an LDL greater than 100) and finally, ACE inhibitors. (Figure 9).^1,24 It is important to understand that hospitalization for an acute event is the ideal opportunity to initiate patients on the appropriate regimen for long-term secondary prevention. The guideline committee noted that the MIRACL trial showed evidence of benefit with lipid-lowering therapy in terms of recurrent ischemia rates.²⁵ However, there is also stronger evidence from the multiple long-term trials that initiating statins at a longer time post-event is beneficial.^26,27 There is also evidence that starting the lipid-lowering agent in the hospital makes the patient much more likely to be on lipid-lowering therapy one year later.²⁸ Thus it is important to place patients on appropriate long-term secondary prevention at the time they are discharged from the hospital. Critical Pathways. Sid Smith could not be here today, but he has stated that critical pathways are necessary because a real gap exists between what practitioners think they know and what they are actually doing in practice. An interesting survey conducted by a colleague at the Brigham and Women’s Hospital questioned local physicians about their NCEP III Guidelines knowledge. Naturally, everyone scored very well, as many people in this profession do on standardized tests.29 The survey revealed, however, that 18% of these outpatients achieved the NCEP III goal of LDL 29 It is evident from this small survey that a huge disparity exists between what is known and what is actually being done in practice. Critical pathway guidelines are designed to narrow this gap. Several well conducted studies show that the application of critical pathways is an effective strategy. The CHAMP program, for example, involves having a nurse use a checklist to ensure that all patients are on appropriate guideline-recommended therapies. Figure 11 shows that aspirin and beta-blocker use, both acutely and at follow-up, was dramatically increased with the CHAMP program.³⁰ Aspirin use was just 68% before the program was launched, but was maintained at 94% one year later. The same increase in use was achieved with beta-blockers, ACE inhibitors, and an enormous improvement was achieved in statin use. This again reinforces what the guidelines put forth: the acute treatment phase presents the ideal opportunity to consider a patient’s long-term treatment strategy and to ensure that he is on the appropriate long-term secondary prevention regimen. Gregg Fonarow was also able to update his critical pathways with the new guideline recommendations, including early and long-term use of clopidogrel.³⁰ He found that a dramatic increase occurred at UCLA in the early- and long-term use of clopidogrel when updated standardized pathways based on the CURE study were employed. This shows that new therapies can get incorporated into clinical practice by using critical pathways. Also of interest is the GAP program, implemented by Kim Eagle and colleagues, which provides data supporting the efficacy of critical pathways.³¹ GAP targeted myocardial infarction patients, looked at early aspirin and beta-blocker use, and the measurement of LDL cholesterol (Figure 12).³¹ Patient charts were examined to ascertain whether the checklist was completed or the standardized orders were used. The cases in which the checklists were completed and the pathway tools were applied showed a boost in the use of aspirin and beta-blockers, reinforcing the notion that physician and nursing aids really can put scientific evidence into practice. Finally, the Get with the Guidelines program offers a web-based tool in which the user can enter data and utilize interactive pop-up menus. This program has reported initial pilot data showing improvements in use of guideline-recommended therapies, and is now expanding nationwide. A final part of the process of improving guideline compliance involves monitoring performance with registry data. Numerous registries in the U.S. and worldwide show practitioners precisely what they are doing. Underutilization of these registry data, even those from the most recent CRUSADE registry, is still quite prevalent.32 The overall message is that gaps clearly exist in our knowledge and critical pathways can help us implement the guidelines needed to close these gaps. DISCUSSION Mitch Driesman: Shamir, could you explain the discrepancy between the guidelines and the data you presented? To my recollection, the guidelines for cardiac surgery recommend stopping clopidogrel use five days prior to elective cardiac surgery, yet the data you presented argue strongly that we should not stop the drug. Were those pre-PCI data? Shamir Mehta: You used the correct term: "prior to elective cardiac surgery." A trend was observed in the major or life-threatening bleeding (p-value 0.06), which may explain the discrepancy between the two. Mitch Driesman: On the other hand, you showed us that there was a 1/1000 life-threatening bleed rate, yet 22/1,000 cardiac events are prevented if clopidogrel is continued through the surgical period. Was the clopidogrel therapy implemented over nine months? Did you observe the benefit in the ten days surrounding the surgery, or over nine months, including surgery and beyond? Shamir Mehta: The benefit occurred in acute treatment with clopidogrel within the first thirty days. Mitch Driesman: I will act as the devil’s advocate here. Your data suggest that we really should not stop clopidogrel use for five days. Shamir Mehta: There is a risk, when the antiplatelet drug is withdrawn from a “hot” patient, that an event could occur between the time the drug is stopped and the time of surgery. We don’t yet know what the risk/benefit ratio is for that scenario. Elliot Rapaport: I would like to point out that Dr. Hongo’s study was a non-emergent study, not an ACS study. Patients in Dr. Hongo’s study were excluded if they had recently received GP IIb/IIIa inhibitors. Thus, his was strictly an observational study involving clopidogrel versus no clopidogrel use at the time of routine CABG surgery. Chris Cannon: Yes, but in any observational study there are many confounding factors that are never accounted for in multivariate analysis. The statistics invariably will hold up in multivariate analysis. We should keep in mind that randomized data will always trump observational data. Of course, observational data can be accurate, but randomized data are always needed to support them. Jeffrey Moses: We had this same debate with aspirin 25 years ago, remember? And everyone is still taking aspirin today! James Tcheng: Clopidogrel use in the first nine months is a Class I-b recommendation in the Guidelines. Why is it not a Class I-a recommendation? Shamir Mehta: I’m not on the committee, although I was a Guidelines reviewer. Unfortunately, I gave Dr. Braunwald 40 pages of single-spaced comments on the guidelines, so I quickly became enemy number-one with the committee! I would say that the I-b recommendation category is due to the recognition that one trial was involved, albeit a huge one. As you pointed out, however, the evidence is just as strong for clopidogrel use in the first 30 days as it is for beyond 30 days. Spencer King: Unlike the GP IIb/IIIa inhibitor trials where most of the benefit was observed in patients who were troponin-positive, the CURE trial showed that the benefit was the same at nine months. Do you have data at 30 days comparing troponin-positive and troponin-negative patients? I am interested to know if the beneficial effect was due to prolonged or even short-term therapy. Shamir Mehta: When the CURE trial was launched, troponins were new to the scene, so they hadn’t yet been integrated into regular practice in all the hospitals and studies, but by the middle of the study and toward its completion, everyone had begun using troponins regularly. In the CURE study, we are talking about enzyme positivity at baseline versus no enzyme positivity. A significant benefit was conferred to those patients who had baseline enzyme positivity — whether CK-mB or troponin. We have collected blood samples (stored at -180º C) from over 8,000 patients in the study and plan to conduct a formal troponin substudy from CURE in a core lab. Chris Cannon: Should we then conduct two 6000-patient studies instead of one 12,000-patient study in order to qualify for a 1-a recommendation? James Tcheng: If one large trial shows a statistic of 0.052, then there are essentially two trials, because the criterion of Ron Waksman: Shamir, I would like to return to the issue of clopidogrel treatment prior to surgery. We face the clinical situation on a daily basis where patients come in for possible catheterization. Clopidogrel can be administered in the holding area before the catheterization, or once it is determined that the patient needs PCI. Were there any data from CURE suggesting a difference in timing — not in terms of days, but an hour or so before the PCI procedure? The data you presented showed that clopidogrel’s effects were observed within two hours. However, the issue is a practical one because surgeons don’t like to operate when a patient is on clopidogrel. If clopidogrel loading was delayed in the patients with three-vessel disease, then they could still be sent directly to a tertiary care center for surgery. Are there any data on this? Shamir Mehta: Are you talking about elective or acute coronary syndrome patients? Ron Waksman: Elective patients. Deepak Bhatt: I don’t have any randomized data to address that point, but David Moliterno has put together the one-year TARGET study results stratified by pre-treatment with clopidogrel versus no pre-treatment with clopidogrel. Of course, every patient received a GP IIb/IIIa inhibitor, either abciximab or tirofiban. Interestingly, what emerges at one year — and these data have been submitted for publication — is that there is actually a mortality benefit in those patients who were pre-treated with clopidogrel. Granted, these results come with the caveat that they are non-randomized. Ron Waksman: When were these patients pre-treated? Immediately beforehand, or one hour beforehand? Deepak Bhatt: Actually, they were pre-treated two to six hours beforehand, which does seem to confer a mortality benefit that persists on multivariate analysis, given the caveat. Chris Cannon: If the patient is in the catheterization laboratory, waiting one hour is probably not significant, especially given what Steve just said. But for the acute coronary syndrome patient the day or two before, where the physician holds off on administering the drug, giving it in the emergency room, there is the lost opportunity of preventing events as seen in the 24-hour event rate curves. On the 24-hour curve, the event rate is about 2.5% in the aspirin-alone group, and is about 1.5% at 24 hours in the aspirin plus clopidogrel group. Thus, the opportunity is lost in the high-risk situation where the patient has an acute coronary syndrome and undergoes catheterization. If measured in days, I personally favor early treatment with clopidogrel because benefit is gained at the risk of only some minor bleeding events. Ron Waksman: It is only a one-hour difference. Surgeons at tertiary care centers still want to operate on the three-vessel disease patients. Chris Cannon: I would advise holding off on clopidogrel if the patient is in the pre-catheterization holding suite. Mitch Driesman: But in the real world, it is often not just a matter of one hour. If an elective patient has a positive stress test, the physician starts him on 300 mg of clopidogrel for the two days prior to the planned procedure. Thus, it is more often a matter of 48 hours. What is the risk versus the benefit in this situation? James Tcheng: In the ESPRIT trial, patients received a 300 mg dose of clopidogrel between 0 to 6 hours prior to the angioplasty procedure. At one year, there was a 20%+ relative risk reduction associated with “pre-treatment.” It is not a question here of days of pre-treatment, but sometimes just pre-treatment on the table. Chris Cannon: With and without a GP IIb/IIIa inhibitor? James Tcheng: The event curves show that the worst outcomes occurred without pre-treatment and without a GP IIb/IIIa inhibitor. The next curve down shows about a 20% relative risk reduction when placebo is compared to pre-treatment with clopidogrel. The next curve down, which is about a 23% relative risk reduction, represents treatment with integrilin only. The best outcomes (40% relative risk reduction) occurred when the two agents were combined. Chris Cannon: So there was an incremental benefit with clopidogrel, even in the presence of integrilin? James Tcheng: Yes, absolutely. There are incremental benefits either way you slice it. Jeffrey Moses: Those results are somewhat like a fusion of the EPISTENT and TARGET studies. Spencer King:In ESPRIT, was the benefit due to the use of clopidogrel after the procedure? James Tcheng: No. All patients received clopidogrel for 28 to 30 days afterward. In fact, every patient received clopidogrel in the ESPRIT trial, it’s just that approximately 30% of the patients were started on the drug after leaving the catheterization laboratory compared to pre-treatment with the drug. Deepak Bhatt: In the TARGET study, all patients received the drug afterward as well. I really do think that the benefit of pre-treatment is heightened in acute coronary syndrome patients. And I think that the benefit — to a lesser degree — extends to elective intervention patients as well. In my view, a few extra pills of clopidogrel pre-treatment is sort of like the poor man’s GP IIb/IIIa inhibitor. Pre-treatment appears to confer an early antiplatelet effect that modulates outcomes in terms of PCI, pre-procedural myocardial infarction, and — some practitioners think — an extended mortality benefit from GP IIb/IIIa inhibition in high-risk patients. Samir Sharma: Jim, do the data you presented suggest that if clopidogrel had been given in the ESPRIT trial approximately two days beforehand, there might not have been such a large difference in outcomes between integrilin versus no integrilin treatment? James Tcheng: No, the data do not suggest that at all. Rather, they suggest that the two therapies are synergistic. Some patients who received the drug a day or so beforehand showed no difference in outcomes, whether they received it on the table or the day before.