Oral glycoprotein IIb/IIIa inhibition—great idea, but it didn’t work

Abstract

With the numerous positive trials of intravenous glycoprotein IIb/IIIa inhibitors (1), it was hoped that the benefit of IIb/IIIa inhibition could be extended to long-term treatment. This class of drugs was designed to target the final common pathway of platelet aggregation, following which platelets form a thrombus. Glycoprotein IIb/IIIa receptor blockers were designed to compete with fibrinogen and von Willebrand factor to prevent these molecules from cross-linking platelets. When high levels of IIb/IIIa receptor blockers were administered in trials, in the range of 80% to 90% of receptors inhibited, clinical events were reduced in patients undergoing percutaneous coronary intervention and in acute coronary syndromes (1–5). A second observation was that all the benefit in reducing events was achieved during the infusion period (4,5). Although these benefits were sustained out to long-term follow-up, there was no added benefit beyond the time of the initial infusion. Thus, it was recognized that for sustained benefit, these drugs would need to be administered for the long term. Accordingly, oral versions of IIb/IIIa receptor blockers were developed, many of which were tested in phase 3 trials beginning in the mid-1990s. Over the past 2 years, the results have been disappointing. In this issue of the Journal, Newby and colleagues summarize the results of the first four trials (6): the OPUS-TIMI 16 (7), the EXCITE (8), and the two SYMPHONY trials (9,10). A fifth trial, BRAVO, was stopped prematurely due to an increase in mortality, and its full results have not been released (11,12). A sixth study (PURPOSE) involving roxifiban (13) was also terminated, and its results are not available. The findings were remarkably consistent across the four trials, despite differences in patient characteristics and trial design; no benefit was seen on the composite endpoint of death, recurrent myocardial infarction, and other recurrent ischemic events, and there was a 31% increase in mortality (6). The authors have gone to great lengths to understand these findings. They determined that trial design was an unlikely explanation, because the results were consistently negative, and neither was the lack of concomitant aspirin use with use of an oral IIb/ IIIa inhibitor another explanation, because worse outcomes were seen with and without aspirin. One leading explanation is the variability in the pharmacokinetics and pharmacodynamics of the orally administered IIb/IIIa inhibitors. As has been described in phase 2 dose-ranging trials of platelet aggregation, the level of platelet inhibition varies widely from patient to patient and within an individual patient during the dosing interval (14,15). Thus, some patients may have levels of inhibition and platelet aggregation as low as zero, whereas others may have levels of inhibition approaching 100%. With such variability, it is not difficult to understand why there would be a lack of a consistent benefit in patients treated with these oral agents at the dosing strategies used. Indeed, the lessons from recent trials have reiterated the importance of a high, steady level of platelet inhibition. When comparing these results with those of eptifibatide in the IMPACT-II study, where the dosing was associated with an inhibition of 50% to 60% (16), there was only a modest reduction in events (17). In contrast, when a higher dose was used in the ESPRIT trial , targeting 90% to 95% platelet inhibition, the benefit was nearly doubled (3). Similarly, the superiority of abciximab over tirofiban in reducing early clinical events has been attributed to the low level of platelet inhibition achieved by tirofiban at the dose used in the TARGET trial (18,19). However, high levels of inhibition have been associated with increased bleeding (14), as seen with the oral IIb/IIIa inhibitors (7–10), suggesting that drugs in this potent class need to be carefully titrated to optimize outcomes. In addition, the high level of inhibition may preclude long-term treatment because of the likelihood of chronic bleeding. A similar perplexing area has been that oral IIb/IIIa inhibitors may have a prothrombotic tendency (7,10). Substudies in OPUS-TIMI 16 have documented increases in P-selectin and CD63, both markers of platelet activation (20,21), whereas in vitro studies have shown that binding of a IIb/IIIa receptor blocker and then dissociation of this blocker from the receptor can leave the IIb/ IIIa receptor open to binding from fibrinogen, which could lead to an increase in platelet aggregation, especially when levels of IIb/IIIa inhibitors are low (22). Because dosing varies among the oral agents, this hypothesis has become an attractive explanation for why increased thrombotic deaths were seen with several but not all of the oral IIb/IIIa inhibitors. Some agents, notably orbofiAm J Med. 2002;112:673– 675. From the Cardiovascular Division, Department of Medicine, Brigham and Women’s Hospital and Harvard Medical School, Boston, Massachusetts. Requests for reprints should be addressed to Christopher P. Cannon, MD, Cardiovascular Division, Brigham and Women’s Hospital, 75 Francis Street, Boston, Massachusetts 02115.