Low-Dose vs Standard-Dose Unfractionated Heparin for Percutaneous Coronary Intervention in Acute Coronary Syndromes Treated With Fondaparinux: Title and subTitle BreakThe FUTURA/OASIS-8 Randomized Trial FREE

In recent years, important advances have been made in the management of non–ST-segment elevation acute coronary syndromes, with new antithrombotic agents, increasing use of percutaneous coronary intervention (PCI) for high-risk cases, and improved secondary prevention. Among newer antithrombotics, fondaparinux, a synthetic factor Xa inhibitor, has recently emerged as an attractive option. Fondaparinux was compared with subcutaneous enoxaparin in the Organization to Assess Strategies in Acute Ischemic Syndromes (OASIS-5) trial, a large international randomized trial.¹ ST elevation acute coronary syndrome was evaluated in the OASIS-6 trial.² In OASIS-5, fondaparinux was noninferior to enoxaparin for the primary efficacy outcome of death, myocardial infarction (MI), and refractory ischemia but halved major bleeding leading to a significant mortality reduction with fondaparinux compared with enoxaparin. However, in OASIS-5, there was a small but significant increase in catheter-related thromboses with fondaparinux in patients who underwent cardiac catheterization or PCI. This result prompted the OASIS investigators and guideline committees to recommend the use of unfractionated heparin as adjunctive therapy at the time of PCI for patients with non-ST-segment elevation acute coronary syndromes who were treated with fondaparinux and undergoing PCI^{3 - 5} ; although the range of dosing recommended differs between the European Society of Cardiology guidelines (50-100 U/kg) and American College of Cardiology–American Heart Association guidelines (50-60 U/kg irrespective of glycoprotein IIb/IIIa [GpIIb-IIIa] use), reflecting uncertainty because the dosing is based on analysis of limited retrospective data.

Independent of the use of fondaparinux, the optimal adjunctive unfractionated heparin regimen to be used during PCI for acute coronary syndromes has not been evaluated. There are wide variations in the doses of adjunctive unfractionated heparin used and continuing uncertainty as to whether measurement of activated clotting time (ACT) to guide dosing is useful.⁶ The optimal dosing of unfractionated heparin should maintain the safety profile of fondaparinux but achieve adequate antithrombin effect to prevent catheter thrombus.

The Fondaparinux Trial With Unfractionated Heprin During Revascularization in Acute Coronary Syndromes (FUTURA)/OASIS-8 trial evaluated the safety of 2-dose regimens of adjunctive intravenous unfractionated heparin during PCI in high-risk patients with non–ST segment elevation acute coronary syndromes initially treated with subcutaneous fondaparinux and referred for early coronary angiography.

The methods for the FUTURA/OASIS-8 study will be published at a later date (P. G. Steg, MD, unpublished data, August 2010). Briefly, this study involved a prospective cohort of high-risk patients presenting to the hospital with unstable angina/non–ST-segment elevation MI treated with subcutaneous fondaparinux and referred for early coronary angiography. Within this cohort, patients scheduled for PCI were eligible for randomization into a double-blind randomized parallel-group trial evaluating adjunctive standard vs fixed low-dose intravenous unfractionated heparin (Figure 1). At each participating site, the study was reviewed and approved by ethical committees or institutional review boards in accordance with national and international regulations. Each participant provided written informed consent before enrollment.

Patients with all of the following criteria were eligible: (1) history consistent with new, or worsening ischemia, occurring at rest or with minimal activity; (2) enrollment within 48 hours of the most recent symptoms; (3) planned coronary angiography, with PCI if indicated, within 72 hours; (4) at least 2 of the following criteria: aged 60 years or older, troponin T or I or creatine kinase MB above the upper limit of normal; electrocardiograph changes compatible with ischemia, ie, ST depression of 1 mm or greater in 2 contiguous leads, T-wave inversion more than 3 mm, or any dynamic ST shifts; and (5) written informed consent.

The exclusion criteria were (1) younger than 21 years; (2) contraindications to unfractionated heparin or fondaparinux; (3) contraindications for angiography; (4) patients requiring urgent (<120 minutes) coronary angiography due to refractory or recurrent angina associated with dynamic ST changes, heart failure, life-threatening arrhythmias, or hemodynamic instability; (5) treatment with other injectable anticoagulants for the qualifying event, unless the last administered dose was 8 hours or more for low-molecular-weight heparins, 60 minutes or more for bivalirudin, 90 minutes or more for unfractionated heparin; (6) hemorrhagic stroke within 12 months; (7) indication for anticoagulation other than acute coronary syndromes; (8) women who were pregnant or breastfeeding or of childbearing potential not using contraception; (9) life expectancy less than 6 months; (10) receiving an experimental pharmacological agent; (11) revascularization procedure for the qualifying event already performed; and (12) creatinine clearance less than 20 mL/min.

Eligible patients were enrolled in the cohort study and received 2.5 mg of open-label fondaparinux subcutaneously once daily. It was recommended that angiography be conducted within a time frame to ensure that same sitting PCI, if performed, was within the 24-hour dosing window of fondaparinux. Following angiography and confirmation that the patient was to undergo PCI, he/she was randomized using a central interactive voice recognition system, from a computer-generated randomization list, blocked by center, and stratified by intent to use GpIIb-IIIa inhibitor use.

Standard-dose regimen using guideline-recommended dose with adjustment to target ACT, the dose of unfractionated heparin was an 85-U/kg bolus with an additional bolus if needed to achieve ACT of 300 to 350 seconds (using the Hemochron or Hemochron Jr device, International Technidyne Corp, Edison, New Jersey) or 250 to 300 seconds (using the Hemotech, Medtronic Hemotec, Englewood, Colorado, device). For patients with planned GpIIb-IIIa inhibitors, the dose was 60 U/kg with an additional bolus to achieve a target ACT of 200 seconds or longer (using either device). Activated clotting time measurements were performed 5 minutes after the initial bolus.

All patients with a fixed low-dose with no adjustment on ACT received a 50-U/kg unfractionated heparin bolus, irrespective of planned GpIIb-IIIa use.

In all cases, the study blind was maintained, using the unblinded study coordinator. For all patients, the maximum bolus dose initially administered was 10 000 U. Up to 2 additional bolus doses could be administered according to an ACT dosing algorithm. Randomized unfractionated heparin was to be given at least 1 minute before insertion of the guidewire, with flushing of the intravenous line or arterial sheath with saline. The ACT was measured in all patients by an unblinded study coordinator; however, investigators were blinded to ACT results. Sham boluses were administered in the fixed low-dose group. In the event of a PCI procedure lasting more than 1 hour, an additional bolus of unfractionated heparin was allowed (guided by ACT in the standard-dose group and a bolus of 40 U/kg in the low-dose group). Patients did not receive direct thrombin inhibitors, low-molecular-weight heparins, oral anticoagulants, fibrinolytic agents, or dextrans. Nonsteroidal anti-inflammatory drugs were discouraged.

The management of patients undergoing angiography and PCI followed the investigator's usual practice and local guidelines, including choice of arterial access site, equipment, and use of GpIIb-IIIa inhibitors. Details are provided (see the eMethods). Cardiac biomarkers were measured routinely prior to PCI and at 6 and 12 hours after PCI. In the event of a repeat or second PCI, open-label unfractionated heparin was administered as adjunctive therapy in accordance with usual local practice. All patients had a 30-day follow-up visit.

The primary outcome was the composite of peri-PCI major bleeding, minor bleeding, or major vascular access-site complications. Peri-PCI was defined from randomization up to 48 hours after the PCI procedure. Major vascular access site complications include large hematoma, pseudoaneurysm requiring treatment, arteriovenous fistula, or other vascular procedures related to the access site.

The assessment of net clinical benefit (key secondary outcome) was based on the composite of peri-PCI major bleeding with death, MI, or target vessel revascularization at day 30. Other secondary outcomes were major and minor bleeding assessed separately; major vascular access-site complications; the composite of death, MI, target vessel revascularization; and the components assessed separately; major PCI-related procedural complications, defined as abrupt vessel closure, new angiographic filling defect representing either angiographic thrombus or major dissection with reduced flow, no-reflow phenomenon, or catheter-related thrombosis (catheter thrombosis events were adjudicated); stroke; definite and probable stent thrombosis, as per the Academic Research Consortium definitions⁷ ; all primary and secondary outcomes except major vascular access complications; and PCI-related procedural complications, but including catheter thromboses, were adjudicated by a blinded central adjudication committee. Definitions of outcomes (eAppendix) were consistent with those previously used in OASIS-5. For those events that were not adjudicated, precise definitions were provided to the investigators to ensure reporting consistency.

The primary outcome (composite of peri-PCI major or minor bleeding and major vascular access-site complications) was assessed using logistic regression investigating the null hypothesis that there is no difference between the groups in the incidence of primary outcome. The logistic regression model used randomized treatment with prerandomization planned GpIIb-IIIa inhibitor use as a covariate. The same model was used for all secondary outcomes. A significance level of .05 with a 2-sided test was used and SAS version 9.1 (SAS Institute Inc, Cary, North Carolina) was used for all analyses.

All outcomes were assessed at 48 hours (defined as peri-PCI) and at day 30, with the exception of major vascular access-site complications and major PCI-related procedural complications which were only assessed during the peri-PCI period. For the time-to-event analyses, a Cox proportional model with stratification for planned GpIIb-IIIa inhibitor use was used to generate hazard ratios and 95% confidence intervals (CIs). The cumulative hazards were estimated using the Kaplan-Meier method with Splus version 8.1 (Insightful Corp, Seattle, Washington).

Six prespecified subgroups (age, sex, body mass index, creatinine clearance, access site for PCI, and planned use of GpIIb-IIIa blockers) were examined for consistency of the primary and key secondary end points. For these subgroups, homogeneity of the odds ratios (ORs) between the different subcategories was tested using the Breslow-Day test for the primary outcome.

Power calculations were derived from the observations of the OASIS-5 PCI subset. Given 1000 patients per treatment group and an estimated 5% incidence for the primary outcome in the population receiving the standard unfractionated heparin regimen, the study had 81% power to detect a 50% relative reduction in the primary end point of bleeding (a magnitude of effect observed with fondaparinux in OASIS-5) between the standard-dose and the low-dose heparin regimens, using a 2-sided 5% significance level. For the key secondary outcome, assuming an estimated 9% incidence in the control group, the study had 80% power to detect a risk reduction of 37.5% between groups.

The FUTURA/OASIS-8 trial was designed to enroll patients who had similar characteristics as those of the subset of patients who underwent PCI in OASIS-5.^{1 ,5} We prespecified that the FUTURA/OASIS-8 trial would compare the major bleeding peri-PCI that was associated with the addition of unfractionated heparin to fondaparinux with the major bleeding observed among patients in the OASIS-5 trial who were randomized to receive fondaparinux or enoxaparin and who had undergone PCI. The 95% CIs of the major bleeding rates observed for the current study were compared with the point estimates observed in the historical OASIS-5 control population who had undergone PCI during their index admission and the first 8 days in the hospital. Both unfractionated heparin regimens were investigated separately.

To determine whether either the unfractionated heparin–dose regimens plus fondaparinux would not increase peri-PCI major bleeding more than what the historical OASIS-5 control population experienced would depend on whether the major bleeding rates observed in the OASIS-5 subset analysis fell within the 95% CIs that were constructed for the current study for both unfractionated heparin–dose groups. The exact 95% CIs are used for the unadjusted proportions. A sensitivity analysis for the historical comparison was performed by adjusting the rates and 95% CIs of peri-PCI major bleeding with logistic regression of both dose groups in the FUTURA trial for differences in baseline characteristics with the OASIS-5 PCI group. The following variables were used for adjustment: age, sex, body mass index, creatinine clearance, diabetes, use of GpIIb-IIIa inhibitors, and arterial access site.

The study was coordinated at the Population Health Research Institute (PHRI), at McMaster University and Hamilton Health Sciences, Hamilton, Ontario. All analyses were performed at the PHRI, independent of the sponsor.

From February 2009 to March 2010, 3235 patients were enrolled from 179 hospitals in 18 countries (eTable 1). Of these, 3156 (97.6%) underwent coronary angiography within 72 hours. Of the 2026 undergoing PCI 1002 were randomly assigned to the standard ACT-adjusted unfractionated heparin regimen and 1024 were assigned to the experimental fixed low-dose regimen (Figure 1). Follow-up was available for all patients at 48 hours and all patients except 2 in the standard-dose group at 30 days.

The patients' baseline characteristics are described in Table 1 and were balanced between groups. Approximately three-fourths of the patients had non–ST-segment elevation MI and one-fourth unstable angina as the initial diagnosis. The median delay from symptom onset to enrollment into the cohort was 19 hours (interquartile range [IQR], 9-29) and the delay from symptom onset to PCI was 27 hours (IQR, 16-42). The median duration of fondaparinux was 3 days. A third of the patients had received commercially available fondaparinux prior to enrollment. The majority of patients received fondaparinux treatment after PCI for a median duration of 2 days (Table 2).

Procedural characteristics are summarized in Table 2. The median amounts of unfractionated heparin administered were 3800 (50 U/kg) and 6400 U (85 U/kg) in the low-dose and standard-dose groups, respectively. In the standard-dose ACT-guided group, 205 patients (20.5%) required an additional dose of unfractionated heparin to reach the target ACT at the start of the procedure. The primary composite outcome occurred in 4.7% of the patients in the low-dose group and 5.8% in the standard-dose group, (odds ratio [OR], 0.80; 95% CI, 0.54-1.19; P = .27). There was a nonsignificant increase in the key secondary outcome (peri-PCI major bleeding, death, MI, and target vessel revascularization at 30 days) in the low-dose group: 5.8% vs 3.9% in the standard-dose group (OR, 1.51; 95% CI, 1.00-2.28; P = .05).

Secondary outcomes are summarized in Table 3. There was no significant difference in peri-PCI major bleeding in the low- vs standard-dose groups (1.4% vs 1.2%; OR, 1.14; 95% CI, 0.53-2.49; P = .73; Figure 2). Peri-PCI minor bleeding was lower in the low- vs standard-dose group (0.7% vs 1.7%; OR, 0.40; 95% CI, 0.16-0.97; P = .04). The major bleeding events up to 30 days were similar between the low- and standard-dose groups (2.2% vs 1.8%; OR, 1.20; 95% CI, 0.64-2.25; P = .57 for logistic regression and HR, 1.20; 95% CI, 0.64-2.23 for Cox proportional hazard model; Figure 2).

The secondary outcome of death, MI, or /target vessel revascularization was similar, although nominally higher, in patients receiving low-dose vs standard-dose unfractionated heparin (4.5% vs 2.9%; OR, 1.58; 95% CI, 0.98-2.53; P = .06 for logistic regression and HR, 1.56; 95% CI, 0.98-2.48 for Cox proportional hazard model). In the subgroup of patients without planned GpIIb-IIIa, the difference was greater: 4.9% vs 2.6%, respectively; OR, 1.90; 95% CI, 1.10-3.30 (P = .02; P for interaction = .17).

The post hoc exploratory composite end points are presented in Table 3. Thrombotic events were not significantly different between the treatment groups. Catheter thrombosis was seen in 5 cases (0.5%) in the low-dose group and 1 case (0.1%) in the standard-dose groups (OR, 4.91; 95% CI, 0.57-42.1; P = .15).

The effect on the primary outcome (Figure 3) and the effect on the composite end points of death, MI, and target vessel revascularization (Figure 4) was explored consistently across the 6 prespecified subgroups analyzed. In addition, no interaction existed with either prior antithrombotic therapy (unfractionated heparin, low-molecular-weight heparin or commercially available fondaparinux) or use of fondaparinux after PCI.

The main population characteristics from the OASIS-5 PCI subgroup and FUTURA are presented in eTable 2. The peri-PCI major bleeding rate (within 48 hours) observed in the fondaparinux group of OASIS-5 PCI (1.5%) was not significantly different from that of FUTURA because it fell between the 95% CI boundaries of the rates observed in FUTURA (standard-dose crude rate, 1.2%; 95% CI, 0.6%-2.1%; adjusted rate, 1.1%; 95% CI, 0.6%-2.1% and low-dose crude rate, 1.4%; 95% CI, 0.7%-2.3%; adjusted rate, 1.2%; 95% CI, 0.6%-2.2%). However, the peri-PCI major bleeding rate of the enoxaparin group (3.6%,) of OASIS-5 exceeded the upper limit of the 95% CIs for both unfractionated heparin regimens and therefore was higher than those observed in both of the FUTURA study groups.

To our knowledge, FUTURA/OASIS-8 is the first large trial to determine the optimal dosage of heparin for percutaneous intervention and the first to compare a regimen not guided by measurement of the ACT to the conventional ACT-guided dosing. The major finding is that low fixed-dose heparin is not superior to standard ACT-guided heparin dosing (on a background of fondaparinux) in terms of preventing peri-PCI major bleeding or major vascular access-site complications. Catheter-related thrombosis was rare in both groups and very rare (0.1%) when using the standard-dose, guideline-recommended regimen of anticoagulation^{8 - 9} on a foundation of fondaparinux. Based on a comparison with the fondaparinux group of the OASIS-5 PCI population,⁵ the addition of either dose of unfractionated heparin to fondaparinux does not increase the rate of major bleeding.

The finding that adding ACT-guided unfractionated heparin to fondaparinux while treating patients with acute coronary syndromes does not increase major bleeding is important in the context of modern PCI practice. Reducing bleeding is potentially important because several studies have suggested that moderate reductions in bleeding may lead to a reduction in longer-term ischemic events, particularly mortality.^{1 ,10} In our study, there was no clinical benefit to using the experimental low-dose regimen, except for a reduction in minor bleeding alone (but not in the combination of major and minor bleeding). The standard-dose regimen was easy to implement with approximately 80% of patients achieving the target ACT with the initial bolus. These findings support using the currently recommended standard ACT-guided dose of unfractionated heparin when performing PCI in patients with non–ST-segment elevation acute coronary syndromes who are treated with fondaparinux.

The OASIS-5 trial demonstrated that fondaparinux, compared with enoxaparin for treatment of non–ST-segment elevation acute coronary syndromes, reduced major bleeding by 48% at 9 days, which was associated with a 17% mortality reduction at 30 days.¹ In patients undergoing PCI, however, the rate of catheter thrombus was higher in those treated with fondaparinux than with enoxaparin (0.9% vs 0.4%, respectively) that has limited the widespread adoption of fondaparinux for acute coronary syndromes management. From a combined analysis of the OASIS-5 and 6 trials, based on 306 patients who received open-label unfractionated heparin while receiving fondaparinux, the catheter thrombus rate was 0.3%.¹¹ This led to the recommendation of additional unfractionated heparin during PCI in patients treated with fondaparinux but this was based on a small number of patients and there was uncertainty about the optimal dose.^{3 - 5} The FUTURA trial now demonstrates in more than 2000 patients that unfractionated heparin during PCI on a background of fondaparinux is associated with a very low rate of catheter thrombus (0.5% in the low-dose and 0.1% in the standard-dose unfractionated heparin groups). Moreover, the addition of unfractionated heparin to fondaparinux in the FUTURA trial did not increase major bleeding when compared with the historical control fondaparinux group of the OASIS-5 PCI population and was even lower than the enoxaparin group of the OASIS-5 PCI population. As a result, using adjunctive unfractionated heparin during PCI for patients treated with fondaparinux prevents catheter thrombosis (the primary downside of using fondaparinux alone) and maintains the key advantage of fondaparinux of a low rate of major bleeding.

Catheter thrombus has not been systematically recorded in many acute coronary syndromes/PCI trials and so the background rate on unfractionated heparin is uncertain. The STEEPLE trial, which compared enoxaparin vs unfractionated heparin (50-70 I/kg with GpIIb-IIIa and 70-100 U/kg without) during elective PCI, reported catheter thrombus rates of 0.13% on enoxaparin and 0.33% on unfractionated heparin.¹² These rates are comparable with both heparin-dose groups in the FUTURA trial.

There have been no randomized trials of heparin-dosing during PCI for acute coronary syndromes in the stent era; however, observational analyses suggest that higher weight-adjusted heparin doses or higher ACT values are associated with increased rates of bleeding.⁶ In patients undergoing balloon angioplasty without stents, there have been 4 published randomized trials^{13 - 16} of heparin dosing during PCI, totaling less than 3000 patients. These trials consistently found that lowering the dose of heparin reduced minor, but not major, bleeding rates, which is consistent with the results of the FUTURA trial. However, a criticism of these previous randomized trials is that they were relatively underpowered for ischemic event rates and may not be applicable to patients undergoing coronary stenting.

Observational analyses in the stent era have shown conflicting results regarding the relationship of activated clotting time and ischemic outcomes.^{6 ,12 ,17 - 18} In a recent international observational study,¹⁹ the average dose of unfractionated heparin used for PCI was 108 U/kg, which indicates that the majority of operators adhere to the guidelines. The large number of institutions and countries participating in FUTURA provide some reassurance that the results are externally valid. Given the paucity of randomized data regarding heparin dosing, further larger studies evaluating the optimal dose of unfractionated heparin are necessary to optimize anticoagulation regimens for PCI and to ensure that reductions in bleeding rates are not achieved at the expense of increased thrombotic event rates.

Despite being the largest randomized trial to assess dosing of heparin for PCI in the stenting era, FUTURA is still underpowered to conclusively rule out moderate, but important, reductions in bleeding from the use of low-dose unfractionated heparin. Based on the observed 5.8% event rate of the primary end point, a sample size of 11 542 patients would be needed to have 80% power to detect a 20% relative risk reduction. In addition, the study was not powered to fully compare the effects of both regimens on ischemic events alone, which precludes a definite assessment of the balance between bleeding and thrombotic risks from each regimen. The design also did not allow for the assessment of whether ACT-guidance or heparin dosing was more important for preventing thrombosis and bleeding because both factors differed between treatment groups. Nevertheless, our study emphasizes that the guideline-recommended practice of using the higher dose of unfractionated heparin with ACT guidance is preferred until more conclusive data on the benefits and risks of lower fixed-dose unfractionated heparin regimen emerges.

Even though the importance and benefits of using unfractionated heparin for some PCI procedures has been questioned in the modern era of routine stenting and dual antiplatelet therapy, at least in the elective setting,²⁰ the FUTURA results emphasize that anticoagulation regimens during PCI remain important because they impact not only on bleeding event rates but also on important clinical outcomes. The historical comparison with OASIS-5 should be interpreted with caution given differences in event reporting and data collection between the 2 studies, unrecognized or unknown factors that cannot be included in the adjustment and the open-label design of OASIS-8 for fondaparinux.

Low-dose compared with ACT-guided standard-dose heparin did not reduce peri-PCI bleeding and vascular access site complications. Catheter thromboses are rare when using unfractionated heparin for PCI in patients with non–ST-segment elevation acute coronary syndromes treated with fondaparinux. Therefore, patients with acute coronary syndromes treated with fondaparinux and undergoing PCI should receive the guideline-recommended ACT-guided standard dose of unfractionated heparin.