Optimal Timing for Use of Glycoprotein IIb/IIIa Inhibitors in Acute Coronary Syndromes: Title and subTitle BreakQuestions, Answers, and More Questions

Contemporary management of acute coronary syndromes (ACS) has advanced in the past decade because of synergies between the tenets of basic science and findings from clinical trials.¹ Clinicians, clinical investigators, and patients depend on definitive evidence to promote changes in clinical care. The current challenge is to create a clinical research system that responds to changes in the science with rapid and efficient planning of studies, acquisition of data, analysis, and dissemination of information. Such a system can provide the foundation for continued improvement in clinical care.²

Large outcome studies are needed to delineate the effects of novel therapies and treatments, as data from small trials or observational registries may be inadequate and even misleading. Human disease is complex and often incompletely understood; the interaction of pharmacologic therapies and management strategies can be unpredictable, and choices of therapies, drug dosages, and timing of interventions in the delivery of health care may be inconsistent or even improper. Thus, unanticipated benefits or risks can only be accurately assessed in large, representative populations.

The ACUITY Timing Trial reported by Stone and colleagues³ in this issue of JAMA is a good example of a clinical investigation that can provide evidence for clinical practice as well as a factual framework that addresses several clinical and research issues. Recognizing an opportunity for evaluation of an important clinical question within the framework of the larger trial, the ACUITY Timing Trial was designed to assess whether the deferred, selective use of glycoprotein IIb/IIIa (Gp IIb/IIIa) inhibitors in patients with non–ST-segment elevation ACS presenting with moderate- to high-risk features is noninferior to routine upstream administration of Gp IIb/IIIa inhibitors. This question was tested as a second randomization in 2 of the 3 randomized groups in the main ACUITY trial.⁴ The methods are well described, but several points should be emphasized. First, randomization in the main trial used an open-label design.⁴ Second, the choice of the Gp IIb/IIIa inhibitor as well as the type of heparin (unfractionated heparin or low-molecular-weight heparin) was at the discretion of the treating physician. Third, the timing of coronary angiography, the decision to undergo percutaneous coronary intervention (PCI), and the use of other adjunctive therapies were also left to the discretion of the treating physician, who was aware of the treatments assigned by both randomizations.

The primary results of the ACUITY Timing Trial show that deferred, selective use of Gp IIb/IIIa inhibitors in patients with non–ST-segment elevation ACS treated with an invasive strategy (99% with coronary angiography) produced a non–statistically significant 12% relative increase in ischemic complications (death, myocardial infarction, or unplanned revascularization for ischemia through 30 days) (7.9% vs 7.1%; relative risk [RR], 1.12; 95% confidence interval [CI], 0.97-1.29); this difference did not meet the predefined criterion for noninferiority. Key secondary end points were major non–coronary artery bypass graft (CABG)–related bleeding defined by the ACUITY scale⁵ and net clinical benefit of the ischemic complications and major non–CABG-related bleeding. A statistically significant reduction in major non–CABG-related bleeding was observed with the deferred, selective use of Gp IIb/IIIa (6.1% vs 4.9%; RR, 0.80; 95% CI, 0.67-0.95; P = .009). The net clinical outcome occurred at a similar rate in both treatment strategies (11.7% vs 11.7%; RR, 1.0; 95% CI, 0.89-1.11; P = .93).

Qualitatively, the key findings of the ACUITY Timing Trial are consistent with what is known about the use of Gp IIb/IIIa inhibitors in patients with non–ST-segment elevation ACS treated with an intended aggressive strategy—these agents reduce ischemic complications before and after PCI but are associated with an increased risk of bleeding.^{6 - 8} Should these data fundamentally influence how Gp IIb/IIIa agents are used in clinical practice? In the CRUSADE registry, the median time from hospital presentation to coronary angiography in patients with non–ST-segment elevation ACS was 23.4 hours during weekdays and 46.3 hours on weekends.⁹ Patients in the ACUITY Timing Trial underwent PCI a median of approximately 5 hours after randomization, so the duration of Gp IIb/IIIa inhibition in the routine upstream study group was much shorter than that which would occur in routine clinical practice. This narrow window of treatment difference makes it difficult to conclude that the 2 strategies of drug use have been adequately tested.

The effect of longer duration of therapy prior to coronary angiography on the ischemic and safety outcomes remains unknown, as the authors acknowledge.³ This important and perhaps unanticipated limitation influences how these results might apply to typical clinical practice outside of the ACUITY sites. The ongoing EARLY ACS trial—in which patients in the setting of a “next-day catheterization strategy” are randomized to upstream therapy in one group and therapy at physician discretion during PCI in the other—should provide additional data that may be more representative of contemporary practice.¹⁰ Until those results are available, however, clinicians of patients with median times of 4 to 5 hours from treatment initiation to coronary angiography could apply the present results to their clinical care algorithms.

In the ACUITY Timing Trial, clinical care decisions were left to the discretion of the treating physicians. This is a common and preferred approach that allows for the testing of research questions in a setting reflective of routine practice, but its use does produce some limitations. In ACUITY, the use of antithrombin therapy was open-label and, in the nonbivalirudin group, the physician had the choice to use unfractionated heparin or low-molecular-weight heparin. In addition, the decision about which Gp IIb/IIIa inhibitor to use upstream or during PCI was left to the discretion of the treating physician. Moreover, two thirds of patients were already receiving some antithrombin therapy prior to randomization, and nearly 10% of patients were being treated with a Gp IIb/IIIa inhibitor prior to randomization.

Use of an open-label strategy simplifies study procedures and avoids the complexity of a double-blind, double-dummy system, and inclusion of patients who had already started therapy prior to randomization allows faster recruitment. These are reasonable study decisions in the context of a large trial, but these decisions also introduce potential complexities in data interpretation. For example, recent data suggest a potential effect of prerandomization therapy on patient outcomes and response to randomized therapy.¹¹ Adjusted analyses are needed to better understand the ischemic and bleeding outcomes in patient populations in the trial defined by prerandomization therapy and “consistent therapy.”³

As occurs in the analysis of many large trials, Stone et al also report a series of subgroup analyses in an attempt to shed light on the trial findings. In addition to the “standard” subgroups defined before randomization, the authors include analyses based on postrandomization variables (eg, the time-to-PCI and duration-of-therapy subgroup analyses) that were performed as an attempt to address the issues surrounding the relatively short median duration of therapy.³ The use of multiple subgroups can be overinterpreted,¹² and subgroup analyses defined by postrandomization events must be viewed cautiously because of the potential for multiple confounders and biases that require particular additional statistical analysis.¹³

The use of superiority testing, noninferiority testing, and a net clinical benefit end point in the ACUITY Timing Trial is complex and may be confusing, but the data and statistical approaches are presented clearly and comprehensively. As the benefits of new therapies become more incremental because of the success of established treatments, clinicians need to understand and accept the design of noninferiority and equivalence trials. Particular attention should be paid to understanding issues related to the establishment of the acceptable margin of difference and to how the standard therapy is actually used.^{14 - 15}

The ACUITY investigators report bleeding events through the use of several assessment tools (ACUITY scale, TIMI scale, and transfusions). Although the relative importance of various bleeding scales has been questioned,¹⁶ the trial results show a consistent increase in bleeding (regardless of how it is defined) with the routine upstream strategy. More data regarding these bleeding events are needed to put these results in proper context. For example, information about the actual doses of anticoagulants, adjustment for renal dysfunction, and possible medication interactions, especially among the elderly, may help to identify risk factors for bleeding. Alexander et al¹⁷ reported that Gp IIb/IIIa inhibitors are frequently overdosed in routine clinical practice and that this excessive dosing is associated with an increased risk of bleeding. This increase in bleeding has to be balanced against a non–statistically significant excess in ischemic complications in the delayed selective strategy that did not meet the criteria for noninferiority (but that could have been as much as 29% worse). Unfortunately, until the risks associated with these events are better clarified, the only prudent recommendation is for clinicians to weigh the risks and benefits of ischemic and bleeding outcomes in their patients on a case-by-case basis. The 1-year follow-up data for patients in the ACUITY trial may provide some valuable insights on the longer-term consequences of these events.

Advances in clinical practice need to be derived from large trials similar in size to the ACUITY Timing Trial, in which 450 centers in 17 countries enrolled more than 9000 patients in slightly more than 2 years. Most recent ACS trials have relied on global site participation and long durations of recruitment. While the global burden of cardiovascular disease is extremely high and patients with acute cardiovascular events are very prevalent, current clinical research systems are inadequate to allow rapid testing of new care strategies. Because millions of patients experience an ACS each year, there is a global imperative to broaden the reach of clinical investigation.

Much discussion has occurred in the academic, industry, and government sectors about the challenges associated with clinical investigation.¹⁸ The time demands of clinical practice, the shortage of cardiovascular specialists, the complexity of regulations and contracts to perform human experimentation, and wavering emphasis on training and mentoring all have contributed to the current situation. In addition, the site-investigator resources required to perform these studies are diminishing. Many practice groups can no longer support clinical research activities. A pivotal challenge for the clinical research community is to align clinical and research goals for seamless integration in routine clinical practice.

Efforts to support new and innovative strategies for the enhancement of clinical investigation that will improve patient outcomes must be promoted. Funding of research institutes through the recently awarded Clinical Translational Science Awards and promising programs through the National Institutes of Health Roadmap, like the Clinical Trials Network (http://www.ctnbestpractices.org), are providing infrastructure for research institutes at academic centers as well as support and guidance for site-based clinical investigators.

Even though the results from the ACUITY Timing Trial should not fundamentally change the use of Gp IIb/IIIa inhibitors in clinical practice, the ACUITY investigators have addressed an important question within the framework of a large clinical trial. Because of varying results from SYNERGY (enoxaparin vs unfractionated heparin),¹⁹ ACUITY (bivalirudin vs unfractionated heparin/low-molecular-weight heparin),³ OASIS (fondaparinux vs enoxaparin),²⁰ and other trials, clinicians currently struggle with what constitutes the optimum antithrombin/antiplatelet strategy for patients with ACS. The ACUITY Timing Trial was not designed to allow comparisons among agents, and these considerations are appropriately absent from this report. Much remains to be learned about optimum care of patients with ACS, and the findings from the ACUITY Timing Trial provide valuable insights for future investigators seeking answers to critical questions of clinical practice.