The choice of anticoagulant in patients undergoing percutaneous coronary intervention (PCI) has been the subject of intense investigation over the past decade. Ever since anticoagulation with heparin was observed to be an essential treatment in patients undergoing balloon angioplasty, unfractionated heparin (UFH) has been the mainstay of procedural anticoagulation. However, use of this therapy is not without side effects, including a high incidence of bleeding when used at standard dose in conjunction with glycoprotein IIb/IIIa inhibitors (GPI).¹ Moreover, activation of platelets occurs with use of UFH, necessitating additional treatment with antiplatelet agents. Two newer anticoagulants, low-molecular weight heparin (LMWH) and bivalirudin have both been extensively studied in the PCI setting. LMWH has greater anti-Xa activity than does UFH, making it a potentially more potent anticoagulant. It also does not activate platelets. However, it has a much longer half-life than UFH, making it a less appealing anticoagulant in the PCI setting. Moreover, there are no commercially available methods to measure anti-Xa activity and the extent of anticoagulation in a patient at the time of PCI, which has limited more widespread use of this class of anticoagulant.

Bivalirudin is a direct thrombin inhibitor, originally derived from the medicinal leech. It has several potential advantages over UFH including the absence of platelet activation. It was evaluated in the landmark PCI trial REPLACE-2², where it was compared to UFH and GPI for PCI in patients with unstable angina. The results of REPLACE-2 indicated that 30-day major adverse cardiac events were comparable between the two treatment groups, and that bleeding rates were significantly lower in the bivalirudin-treated patients. The results of this study paved the way for the introduction of bivalirudin into routine use as an anticoagulant during PCI. The results of the more recent ACUITY study confirms and extends these findings to patients with moderate and high-risk acute coronary syndromes.³ While the benefit of bivalirudin use in general clinical practice has been shown in several studies, almost all of the data were obtained during the era of the bare-metal stent (BMS). It is not clear that there are dramatic differences in the need for anticoagulation between BMS and drug-eluting stents (DES). However, recent concerns about potential differences in the propensity for stent thrombosis between these two stent types makes it reasonable to consider this issue in the DES era.

In this issue, Mishkel et al present the results of a retrospective, single-center experience with the use of bivalirudin and UFH plus GPI in patients who did not have a myocardial infarction (MI), undergoing PCI at the author’s institution.⁴ The authors found a low rate of in-hospital adverse cardiac events for both treatment groups. There was, however, a significantly higher rate of vascular complications in the UFH plus GPI group than in the bivalirudin group. At a mean follow up of almost 2 years, event-free survival was similar in both groups. The authors also performed a cost analysis of the two treatment strategies. The unadjusted and adjusted hospital costs were similar for the two strategies, with the vast majority in the UFH plus GPI group receiving eptifibatide. In the subset of patients receiving abciximab, hospital costs were on average higher than the bivalirudin group by the difference in the cost of abciximab (approximately $1,000).

The results of this study add information to the growing body of literature addressing the use of anticoagulants during PCI in the DES era. The authors conclude that “these observations suggest that bivalirudin is a safe, cost-effective alternative to heparin plus GPI in patients undergoing DES in absence of acute MI.” It would be useful to look more closely at the potential biases in the study that could affect this conclusion. First, there were differences in the baseline characteristics of the two groups that may have affected the study outcomes. There were more unstable angina patients in the UFH plus GPI group than the bivalirudin group. We do not know how many elective patients were treated in this study, nor how many of the unstable angina patients received upstream GPI treatment. Differences in severity of the baseline clinical syndrome may indicate that these were not comparable study groups, so that the meaning of comparable outcomes is problematic. Second, there was a significant difference in the choice of DES —sirolimus-eluting versus paclitaxel-eluting — in the two treatment groups. This suggests that a bias existed within the study among the interventionalists in choice of both anticoagulant and DES type. What effect this potential bias had on study outcomes is also problematic.

These potential limitations of the study notwithstanding, the study results provide reassurance that bivalirudin can be used in conjunction with DES in a group of non-MI patients with a very low rate of procedure-related complications. It has become clear that a gradient of risk is present in PCI patients, with the highest-risk patient benefiting from maximal antiplatelet and anticoagulation therapy, and the lowest-risk requiring simpler, less costly strategies.⁵ Lessons learned from studies such as this help to better define the optimal strategies for PCI as we strive to provide the highest quality and most cost-effective care possible.