Late-Breaking Clinical Trials: American College of Cardiology Scientific Sessions—March 1999
Note: The following reports are summaries and comments on some of the more interesting late-breaking clinical trials presented at the American College of Cardiology Annual Scientific Sessions, March 8-10, 1999, in New Orleans, LA. The reports are based upon handwritten notes taken during the presentation of data and the comments are by the notetakers. No abstracts or written reports were available for these sessions and some of the data presented were preliminary.
The shock trial, an international multicenter study (30 hospitals) enrolling patients with acute myocardial infarction (AMI) and cardiogenic shock between 1993-1998, tested the hypothesis that immediate revascularization would reduce 30-day and 6-month mortality. There is suggestive evidence from observational studies that there may be up to a 30-40% benefit from early revascularization in such patients, although not all data are supportive. Eligible patients with AMI and cardiogenic shock within 36 hours were randomized to either emergency revascularization or supportive therapy with optional late revascularization after a minimum of 54 hours. The hypothesis was that the revascularization group (REV) would have a 20% decrease in mortality compared to the delayed cohort, who had initial medical stabilization (IMS). Patients had sustained hypotension with end-organ malperfusion and no mechanical cause of shock. Wedge pressure had to be more than 15 mmHg and cardiac index less than 2.2 L/min/m2. The IMS group all received intra-aortic balloon (IABP) support and thrombolytic therapy; delayed revascularization was an option, and was carried out in approximately 25% of this cohort.
The overall SHOCK cohort was felt to be representative of an acute infarction-cardiogenic shock population. Fifty-five percent of the patients were transferred from an outlying hospital to one of the participating centers. A total of 302 patients were randomized, representing approximately 25% of eligible candidates.
The primary end point of 30-day all-cause mortality was not achieved, but survival did favor the REV group. Thirty-day mortality was 48% vs. 56%, and six-month mortality was 54% vs. 66%, REV vs. IMS, respectively. Although 30-day mortality rates did not achieve statistical significance, there was a relative risk reduction of 17%, whereas the six-month survival was significantly increased (P = 0.04). There were an equivalent number of major adverse events in both groups, although acute renal failure was more common in the REV patients. A registry was kept of nonenrolled patients and survival results were roughly equivalent. A total of 87% of the REV group actually received angioplasty or bypass surgery (49% and 38%, respectively) after randomization. Of interest, when the cohorts were analyzed with respect to age, there was a favorable trend toward improved survival in those individuals younger than 75 years of age, whereas in patients older than 75 there was an increased risk ratio. Other parameters, such as infarct location, diabetes, or gender, did not affect the outcomes between the two groups.
An angiographic analysis was also reported; approximately 20% had left main disease, and two- or three-vessel disease was found in 80%. Half of the patients had a left anterior decending culprit lesion. TIMI-3 flow was achieved in approximately 60% of the angioplasty patients; there was increasing use of stents in the last years of the trial; stenting did not improve TIMI flow, but did result in increased procedural success rates. Patients who were sent for bypass surgery had more left main and three-vessel disease. In those who underwent a successful angioplasty, 30-day mortality was 38% vs. 79% in subjects with an unsuccessful procedure and 100% in those individuals who did not achieve TIMI-2 or TIMI-3 flow. A separate echocardiographic substudy was reported at another session. An initial echo was performed within 24 hours and a subsequent echo at two weeks or discharge. Many patients were on IABP. Approximately 90 patients in each group were included in this study. One-third did not have their initial echo until after revascularization. Severe mitral regurgitation was an exclusion to enter into the SHOCK study. Echos demonstrated 1-2+ mitral regurgitation in most patients and a tendency for left ventricular volumes to increase over time. The baseline ejection fraction was 33%. Sphericity was increased. Ventricular thrombus was noted in 25% of the IMS group vs. 9% in REV. Left ventricular ejection fraction (EF) did not change significantly over the first two weeks. Individuals with an EF of less than 25% had a 65% 30-day mortality compared to a 30-40% mortality in those with an EF of more than 25%. Left ventricular function tended to improve after revascularization. Mortality was correlated with low EF and mitral regurgitation severity.
Comment by Jonathan Abrams, MD
The investigators in the SHOCK Trial are to be congratulated on carrying out a difficult protocol in extremely ill individuals. While a 20% mortality reduction was not achieved in individuals revascularized within six hours, this may be in part because the intensive medically treated patients received superb care, including IABP and thrombolysis. In addition, approximately 25% of the IMS group underwent revascularization after 54 hours. Thus, the SHOCK Trial actually investigated urgent revascularization vs. optimal medical therapy with IABP. The latter group achieved a high level of care in the study centers, as attested by the relatively good mortality at 12 months. The observation that individuals younger than 75 appear to have a survival benefit is considered to be preliminary until the data are further analyzed; however, a six-month mortality of 48% in the younger cohort is impressive. Furthermore, successful angioplasty was associated with a 38% 30-day mortality; for REV patients who underwent either angioplasty or CABG, the 30-day mortality rate was 42% vs. 53% in those who did not undergo revascularization.
Recent data from the National Registry of Myocardial Infarction-2 (NRMI-2) shock cohort, reflecting 23,000 patients with cardiogenic shock post-AMI, documents an overall mortality of 70%, with or without receiving thrombolysis (Barron HV, personal communication; Barron HV, et al. J Am Coll Cardiol, 1998;31:135A). In patients who received thrombolytic therapy and IABP, mortality was 49%; the investigators concluded that the use of balloon pumping and thrombolytic therapy decreased the odds of death by 18%. In the entire NRMI-2 shock registry, 24% received thrombolytic therapy, 15% underwent early revascularization, and 60% received no immediate reperfusion treatment. IABP was performed in 32%. Those individuals who received IABP tended to be younger, were more likely to be male, and had lower comorbidity than those individuals who did not receive IABP. The hospital mortality rate for patients treated with primary angioplasty with or without IABP was approximately 45% in the NRMI-2 registry, compared to 49% in those individuals who received thrombolytic therapy and IABP. These results are comparable to the 30-day mortality in the SHOCK Trial REV patient.
What conclusion can be drawn from the SHOCK Trial? It seems clear that contemporary therapy of cardiogenic shock complicating AMI demands immediate exclusion of a significant mechanical defect, such as acute severe mitral regurgitation or ventricular septal defect, with initiation of IABP and thrombolytic therapy or direct angioplasty as soon as possible. The SHOCK Trial did not directly test thrombolytic therapy vs. direct angioplasty, but rather an early revascularization strategy vs. stabilization with intra-aortic balloon pumping and thrombolysis with optional revascularization after two or more days. Cardiogenic shock mortality in the SHOCK Trial and the NRMI-2 registry is lower than in previous decades, and convincingly supports a survival improvement in this devastating complication of AMI. One can cautiously conclude that for patients younger than the age of 75, a direct catherization approach followed by angioplasty or bypass surgery is the therapy of choice in suitable patients when treated by highly experienced cardiologists and surgeons. (Editor’s note: More information was recently published in the following article: Goldberg RJ, et al. N Engl J Med 340:1162-1168.)
FRISC II comprises two multicentered trials conducted in Scandinavia: FRagmin during InStability in Coronary artery disease (FRISC) and Fast Revascularization during InStability in Coronary artery disease (FRISC). The first hypothesis tested was that low molecular weight heparin (Fragmin) given bid for three months would reduce death and myocardial infarction rates (primary endpoint). Secondary end points included the need for revascularization and bleeding complications in patients with unstable angina or non-Q-wave myocardial infarction. The second hypothesis tested was the strategy of a noninvasive evaluation vs. an early invasive strategy with revascularization of appropriate patients. Criteria for enrollment in the study were two of the following three: characteristic chest pain (present in 82%), ST-T wave changes (50%), or elevated CK-MB or troponin (60%). Since cardiac catheterization and revascularization was not feasible in all 58 Scandinavian centers in the trial, 2457 patients were in the Fast Revascularization subgroup. During the first five to seven days, all eligible patients were treated with aspirin, beta blockers, nitrates, and Fragmin and those in the strategy subgroup were randomized within 48 hours to either the noninvasive or the invasive arm. The subsequent double-blind phase randomized patients to Fragmin for another three months or placebo. In the strategy subgroup, those in the invasive arm underwent revascularization within the first seven days. Those in the noninvasive arm only underwent revascularization if clinically indicated.
The comparison of low molecular weight heparin vs. placebo showed a small but significant reduction in the primary end point of death or Q-wave myocardial infarction at 45 days (3.7% vs 6.5%; P < 0.003), but at 90 days, this difference was no longer statistically significant (6.7% vs 8.0%). In the strategies evaluation, 98% of the patients randomized to the invasive group underwent angiography and 71% underwent revascularization within 10 days and 78% underwent revascularization by six months. In the noninvasive group, only 9% were revascularized by 10 days and 38% by 6 months. Of note, the mortality in the patients undergoing coronary artery bypass graph surgery was 1.2%.
The primary end point was significantly reduced in the early revascularization group compared to the conservative strategy at six months (9.5% vs 12%; P = 0.045), which represents a 21% reduction in risk. Mortality alone was not significantly different in the total cohort, but it was in the male subjects (1.5% vs 3.2%; P = 0.03). One reason women may not have demonstrated the mortality benefit was that 30% of them had normal coronary arteries on angiography. Bleeding complications on low molecular weight heparin were higher; especially severe bleeding (4.1% vs 1.7%) and cerebral hemorrhage (5 vs 0). No significant change in platelet count was detected on low molecular weight heparin. The authors concluded that low molecular weight heparin produced a modest decrease in the primary end point at 45 days, which was offset by higher bleeding complications and the effect was not present at three months. Thus, they could not strongly recommend prolonged low molecular weight heparin treatment. On the other hand, an early invasive strategy significantly reduced the primary end point at six months, suggesting that for unstable coronary artery syndromes, early invasive management should be recommended with individualization of low molecular weight heparin treatment as clinically indicated. Since the primary end point was reduced at 45 days on low molecular weight heparin, it could be used to help tide patients over until the revascularization strategy could be applied to them.
Comment by Michael H. Crawford, MD
This trial was heartening to the interventional cardiologists who were depressed after the VANQWISH trial showed that the invasive strategy in unstable coronary syndrome patients resulted in higher mortality rates and was not recommended as an initial strategy. The TIMI III trial also showed no advantage to an early invasive strategy. Thus, this is the first trial to show an advantage for the invasive strategy. During the discussion, Lars Wallentin, MD, who presented the data, commented that there were three potential explanations for the differences in these three trials. First, in the FRISC II trial, almost 80% of the patients underwent revascularization in the invasive arm. This was not the case in the two U.S. trials, where it was 44% in the VA trial and about 60% in the TIMI trial. The fact that there were patients in the U.S. trials referred to the invasive arm who did not undergo revascularization suggests that these are different patients than those seen in the FRISC II trial. It is unlikely that they had minimum or no disease since the VANQWISH trial was done in a VA population. Thus, it is more likely that the U.S. patients had more severe three-vessel disease that was not amenable to angioplasty and may not have been particularly attractive for bypass surgery. This impression is supported by Dr. Wallentin’s second point, which was that the U.S. trials’ patients had more comorbidities, such as diabetes. Third, Dr. Wallentin noted that they had a low surgical mortality compared to the U.S. trials, especially the VANQWISH trial. This further supported the contention that the Scandinavian patients were a less sick group in general, and with less extensive coronary artery disease. In support of this was the interesting results in the women patients in the invasive arm, 30% of them had normal coronary arteries.
In the VANQWISH trial, it was not just the higher surgical mortality that contributed to the increased mortality in the invasive arm, since many of the patients died before revascularization could be accomplished. Just performing a cardiac catheterization was associated with a higher mortality in the VANQWISH trial. Also, when the surgical mortality in the VANQWISH trial was adjusted for comorbidities, it was not appreciably more than expected. All these data point to the fact that the U.S. trials dealt with a less favorable patient population. In the United States, we have seen an explosion in the number of patients with non-Q-wave myocardial infarction because of treatment with thrombolytics and more patients who have already been revascularized coming in with ischemic events. Perhaps the de novo unstable angina or non-Q infarction patients, especially if they are younger and less likely to have severe three-vessel disease, should undergo an early invasive strategy and revascularization if feasible. Patients with prior revascularization or known multivessel coronary artery disease should perhaps be treated more conservatively unless they do not do well. In some patients, this treatment might include low molecular weight heparin if catheterization and revascularization decisions are being delayed beyond the first week. In other patients, it may be appropriate to treat with platelet glycoprotein 2B/3A inhibitors. It is of interest that in the FRISC II trial, only 10% of patients receive such agents and this number was equal in all groups. Until the FRISC II study is published and we can carefully look at subgroups, we are going to have to use our best clinical judgment, since the evidence-based medicine approach conflicts with these disparate trials.
The multicenter optimal angioplasty vs. primary stenting study was presented by W. Douglas Weaver, MD. This trial is the first to compare primary stenting with stenting only if plain old balloon angioplasty (POBA) is unsuccessful. The hypothesis tested was that primary stenting would reduce the incidence of death, myocardial infarction (MI), or re-revascularization of the target vessel within six months. OPUS randomized 479 patients with stable or unstable angina, or recent MI. Results showed that significantly fewer patients in the primary stent group reached the primary end point compared to the POBA plus stent if necessary group (6.1% vs 14.9%). Also, a cost analysis was done that showed that the POBA plus or minus stent group cost less initially vs. primary stenting, but when the added cost of readmission for MI or need for re-revascularization was factored in, the overall costs were less in the primary stenting group. Thus, the authors concluded that primary stenting resulted in better clinical outcomes and lower costs than POBA plus stenting if necessary.
Comment by Michael H. Crawford, MD
Previous studies have shown that stents are associated with lower rates of restenosis than POBA, but these differences were not large and primary stenting is a more expensive approach. Several editorials have been written on this topic in major journals, with some discussants decrying "stent-o-mania" over the more reasoned approach of POBA with stenting only if necessary. Others have touted the lower restenosis rate of stent use and, therefore, the potential for lower long-term costs. Another issue in this debate has been the high cost of stents in the United States compared to other countries. Stents and platelet glycoprotein 2B/3A inhibitor drugs have broken the bank at some hospitals and have contributed to increased tensions between hospital administrators and cardiologists. In this milieu, a well-done clinical trial with definitive results in favor of primary stenting is certainly welcome and supports the usual practice of the majority of cardiologists in the United States.
The CLopridigel Aspirin Stent Intervention Cooperative Study (CLASICS) was conducted in 48 European centers and tested the hypothesis that clopridigel plus aspirin would be superior to ticlopidine plus aspirin in patients undergoing coronary artery stenting. Previous studies have shown that ticlopidine plus aspirin reduced in-stent thrombosis, but was associated with significant side effects. Thus, this trial comparing clopridigel to ticlopidine had the primary end point of 28-day safety. Secondary end points were death, myocardial infarction, or need for re-revascularization. This was a randomized double-blind study looking at two doses of clopridigel with standard aspirin therapy vs. aspirin plus ticlopidine. In the high-dose clopridigel group, an intravenous clopridigel load was given first followed by oral therapy. In 90% of the patients, only one stent was placed. The results showed that the primary end point was lower in all clopridigel patients vs. the ticlopidine-treated patients (4.6% vs 9.1%; P < 0.005) with a relative risk reduction of 50%. Patients treated with low-dose clopridigel had a higher event rate than those treated with the intravenous load followed by oral therapy (6.3% vs 2.9%). Bleeding complications occurred in less than 1% of patients and were not statistically different between the groups. Clopridigel did not affect white blood cell counts and there was a minimal decrease of platelets that was of no clinical significance. Also, fewer patients discontinued clopridigel for side effects. The secondary end point of death, infarction, or re-revascularization was not statistically different between the groups. The authors concluded that clopridigel plus aspirin was safer than ticlopidine plus aspirin in patients treated with coronary stenting, a clopridigel intravenous load in the cath lab was well tolerated without increased bleeding risks.
Comment by Michael H. Crawford, MD
Ticlopidine was a significant advance over coumadin treatment to prevent early thrombosis of stented coronary arteries and ticlopidine plus aspirin has been shown to be superior to aspirin alone in preventing in-stent thrombosis. However, ticlopidine is expensive and associated with considerable side effects. Thus, when clopridigel came on the market, there was considerable enthusiasm for its use because of its lower cost and better side effect profile. Thus, many interventional cardiologists have already switched to clopridigel plus aspirin in patients receiving stents. The CLASIC Study supports this choice and also suggests that an intravenous load of clopridigel in the cardiac catheterization laboratory may be the best approach, since there was no increase in bleeding and the primary end point was lowest with this approach. Thus, for reasons of cost and side effects, clopridigel would seem to be the drug of choice in the poststent patient.
Unfortunately, the secondary end point of death, MI, and re-revascularization was not influenced by the various therapies tested. One reason for this is that the event rate was low and the study was underpowered to detect differences in clinical outcome. Thus, we will need to await the results of other trials to determine whether clinical outcomes are better in patients treated with intravenous clopridigel in the catheterization laboratory vs. those just started on oral therapy afterward.