This Article Abstract Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Aguirre, F. V. Articles by Califf, R. M. Search for Related Content PubMed PubMed Citation Articles by Aguirre, F. V. Articles by Califf, R. M. Pubmed/NCBI databases Substance via MeSH Medline Plus Health Information Angioplasty Vascular Diseases

(Circulation. 1995;91:2882-2890.)
© 1995 American Heart Association, Inc.

Articles

Bleeding Complications With the Chimeric Antibody to Platelet Glycoprotein IIb/IIIa Integrin in Patients Undergoing Percutaneous Coronary Intervention

Frank V. Aguirre, MD; Eric J. Topol, MD; James J. Ferguson, MD; Keaven Anderson, PhD; James C. Blankenship, MD; Richard R. Heuser, MD; Kristina Sigmon, MA; Marc Taylor, MD; Ronald Gottlieb, MD; Gary Hanovich, MD; Michael Rosenberg, MD; Thomas J. Donohue, MD; Harlan F. Weisman, MD; Robert M. Califf, MD; for the EPIC Investigators¹

From the St Louis (Mo) University Health Sciences Center (F.V.A., T.J.D.); Cleveland Clinic (Ohio) (E.J.T.); Texas Heart Institute, Houston (J.J.F.); Centocor, Inc, Malvern, Pa (K.A., H.F.W.); Geisinger Medical Center, Danville, Pa (J.C.B.); Arizona Heart Institute and Foundation, Phoenix (R.R.H.); Duke University Medical Center, Durham, NC (K.S., R.M.C.); Deborah Heart Center, Brown Mill, NJ (M.T.); Graduate Hospital, Philadelphia, Pa (R.G.); North Memorial Medical Center, Robbinsdale, Minn (G.H.); and Lutheran General Hospital, Park Ridge, Ill (M.R.).

Correspondence to Frank V. Aguirre, MD, Division of Cardiology, St Louis University Health Sciences Center, 3635 Vista Ave at Grand Blvd, St Louis, MO 63110.

	Abstract

Top Abstract Introduction Methods Results Discussion References

 
Background The potential for novel antiplatelet and antithrombin agents to contribute to periprocedural bleeding complications of percutaneous coronary revascularization is poorly defined. In the Evaluation of c7E3 Fab in Preventing Ischemic Complications of High-Risk Angioplasty (EPIC) trial, the periprocedural use of aspirin, heparin, and a chimeric antibody to the platelet glycoprotein IIb/IIIa integrin c7E3 Fab in 2099 patients significantly reduced postprocedural ischemic complications and 6-month clinical restenosis but was associated with increased procedural bleeding complications. We review these complications and describe clinical and procedural variables associated with increased bleeding complications in the EPIC trial.

Methods and Results Patients with high-risk clinical or lesion morphological characteristics were randomized to receive placebo bolus plus placebo infusion, c7E3 Fab bolus plus placebo infusion, or c7E3 Fab bolus plus c7E3 Fab infusion. Patients received periprocedural aspirin and intravenous heparin continued for a minimum of 12 hours after the procedure. Outcomes reflecting bleeding complications were measured: transfusions, decreased hemoglobin, and an index including both parameters. Major bleeding complications unrelated to bypass surgery occurred in 3.3%, 8.6%, and 10.6%, and blood product transfusions were used in 7.5%, 14.0%, and 16.8% of patients treated with placebo, bolus c7E3 Fab, and bolus plus infusion c7E3 Fab, respectively (both P<.001). Most major bleeding complications occurred at the femoral access site, regardless of treatment. Intracranial hemorrhage (0.3%) and death (0.09%) attributable to major bleeding complications were rare. Multivariable regression analyses identified several variables significantly and independently related to major bleeding complications or greater blood loss, including greater age, female sex, lower weight, c7E3 Fab therapy, and duration and complexity of the index procedure. Major bleeding complications and blood loss in patients receiving bolus plus infusion were not significantly greater than in those receiving bolus alone (P=.38 and P=.14, respectively).

Conclusions Bleeding complications unrelated to bypass surgery were two to three times more frequent in patients receiving c7E3 Fab than in those receiving placebo, but most were transient and well tolerated. Risk-factor analysis and modification of concomitant antithrombotic and antiplatelet treatment strategies may aid in reducing bleeding complications and enhancing clinical benefit in patients receiving c7E3 Fab during percutaneous coronary revascularization.

Key Words: angioplasty • platelet aggregation inhibitors • clinical trials • catheterization

	Introduction

Top Abstract Introduction Methods Results Discussion References

 
Experimental^{1 2} and clinical^{3 4 5} evidence supports a significant role for platelet- and thrombin-mediated mechanisms in the pathogenesis of abrupt vessel closure and acute ischemic complications after percutaneous transluminal coronary revascularization (PTCR). Pharmacological attenuation of platelet- and thrombin-mediated coagulation contributes to reducing post-PTCR ischemic events but also predisposes to a greater risk of hemorrhagic complications.^{6 7 8 9} The increasing procedural complexity of PTCR, together with the adjunctive use of progressively more potent periprocedural antiplatelet and antithrombin regimens, could further increase this risk. The addition of novel antithrombotic and antiplatelet drugs to the standard armamentarium of heparin and aspirin should also have an incremental effect on post-PTCR bleeding complications.

Defining the bleeding risk in PTCR patients, especially those in high-risk subgroups, is clinically important, since the usefulness of newer antithrombin and antiplatelet agents ultimately depends on the demonstration of clinical benefit that overrides any significant increase in risk (ie, bleeding complications). In the Evaluation of c7E3 Fab in Preventing Ischemic Complications of High-Risk Angioplasty (EPIC) trial, the use of a monoclonal chimeric antibody Fab fragment to the glycoprotein (GP) IIb/IIIa platelet receptor (c7E3 Fab), in combination with standard dosing regimens of intravenous heparin and aspirin, was shown to reduce 30-day post-PTCR ischemic events and to limit 6-month clinical restenosis rates in high-risk PTCR patients compared with standard therapy of heparin and aspirin alone.^{10 11} Accordingly, the purpose of the present analysis is to define the frequency of hemorrhagic complications and identify clinical and treatment variables associated with a high risk of bleeding complications among these patients.

	Methods

Top Abstract Introduction Methods Results Discussion References

 
Patient Population
The inclusion and exclusion criteria of the EPIC trial have been described.¹⁰ In brief, 2099 patients <80 years of age undergoing conventional PTCR (ie, balloon coronary angioplasty or directional coronary atherectomy) who had clinical or angiographic characteristics associated with a high risk for abrupt vessel closure were enrolled in the study. These patients included those with acute myocardial infarction of 12 hours' duration undergoing direct or rescue PTCR, with medically refractory unstable or postinfarct angina associated with ECG changes, or with high-risk clinical and angiographic characteristics defined by a modification¹² of the American College of Cardiology/American Heart Association criteria,¹³ alone or in combination with high-risk clinical features such as female sex, diabetes, or age 65 years.

Patients were ineligible for study participation if they had risk factors for bleeding complications. Patients were enrolled in 56 centers in the United States between November 1991 and November 1992. Informed consent was obtained from all patients before study enrollment.

Treatment Protocol
Patients received aspirin (325 mg) beginning 2 hours before PTCR and continuing daily after the procedure. Intravenous (IV) heparin was initiated before PTCR with a bolus of 10 000 to 12 000 U. Subsequent bolus doses (3000 U) were given every 15 minutes during the procedure (total dose, 20 000 U) to achieve an activated clotting time of 300 to 350 seconds. Continuous IV heparin (1000 U/h) was administered for at least 12 hours after PTCR to maintain an activated partial thromboplastin time (aPTT) of 60 seconds or at least twice the baseline value. Further IV heparin therapy was left to the discretion of the individual investigator. Protocol aPTTs were obtained at baseline, at 12 and 36 hours after initiation of therapy, and at discharge.

Randomization and Interventional Procedure
Patients were randomly assigned to one of three treatment groups: placebo bolus plus placebo infusion, c7E3 Fab bolus (0.25 mg/kg) with placebo infusion, or c7E3 Fab bolus (0.25 mg/kg) with a continuous infusion of c7E3 Fab (10 µg/min). The bolus was given within 10 to 60 minutes before the procedure and the infusion for 12 hours unless clinically contraindicated. Chimeric c7E3 Fab is a Fab fragment of a human/mouse hybrid of a murine immunoglobulin molecule that selectively binds to the platelet GP IIb/IIIa receptor (Centocor, Inc).

After PTCR, femoral vascular access sheaths remained in place for at least 6 hours after discontinuation of the study drug infusion and at least 4 hours after discontinuation of the heparin infusion. Sheaths were removed only after an aPTT adequate to achieve hemostasis was obtained (ie, at the physician's discretion). The use of venous sheaths, need for coronary stent placement or intra-aortic balloon pump insertion, procedural use of thrombolytics, and requirements for emergent or elective coronary artery bypass graft surgery (CABG) were mandated according to local practice patterns at each clinical site.

Bleeding Complications
Patients were evaluated daily for overt evidence of clinical bleeding and for hemoglobin loss via laboratory assessment. The classification and site(s) of clinical bleeding were recorded at each site and confirmed by an independent, blinded event committee. Protocol platelet counts were obtained at baseline, at 30 minutes and at 2, 12, and 24 hours after initiation of therapy, then daily until hospital discharge. Protocol hemoglobin and hematocrit concentrations were obtained at baseline, at 12 and 36 hours after the start of infusion, and at hospital discharge.

Bleeding complications were categorized as major or minor according to the Thrombolysis in Myocardial Infarction (TIMI) study group criteria.¹⁴ A major bleeding complication was defined as intracranial hemorrhage, recorded as a new neurological deficit with computerized axial tomographic evidence of blood density; a drop in hemoglobin of >5 g/dL from baseline; or (in the absence of a hemoglobin measurement) a hematocrit decrease of 15% from baseline. Major bleeding complications were defined as occurring either spontaneously in a major organ (eg, intracerebral, gastrointestinal, or genitourinary) or nonspontaneously (eg, provoked by vascular puncture catheters or gastrointestinal tubes, or involvement of a nonmajor organ site).

A minor bleeding complication was defined as (1) spontaneous and observed hematuria or hematemesis, (2) observed blood loss associated with a decrease in hemoglobin of >3 g/dL (hematocrit change, 10%), or (3) a decrease in hemoglobin of 4 g/dL (hematocrit drop, 12%) in the absence of clinical bleeding. All bleeding complications were designated according to time of occurrence into two time intervals: 36 hours or >36 hours after study drug infusion.

An estimate of net blood loss was calculated with a modification of the criteria established by Landefeld et al,¹⁵ by addition of the number of units of packed red blood cells transfused to the baseline-minus-nadir hemoglobin (ie, units transfused plus change in hemoglobin). The study protocol did not include an algorithm for blood product transfusions. All transfusions were prescribed according to the local standard of care at each clinical site.

Statistical Analyses
Continuous demographic and hematologic variables were represented by the median and 25th and 75th percentiles. Comparisons of continuous outcomes were made with nonparametric methods using normal scores; tests for differences among treatment groups were performed with a test for trend from placebo to bolus to bolus plus infusion. Categorical data were described as percentages; pairwise comparisons between treatment groups were made with Fisher's exact test. Tests for differences among all three treatment groups were made with a Cochran-Mantel-Haenszel ² test for trends from placebo to bolus to bolus plus infusion. One-sided confidence bounds for hemorrhagic stroke rates were computed with exact binomial probabilities. Multivariable logistic modeling was applied to examine the relations of major bleeding events to clinical baseline characteristics and to other procedural outcomes. Linear regression modeling was used to test for associations between estimated blood loss and baseline characteristics and between blood loss and procedural outcomes. Statistical testing and variable selection (but not estimation) for the linear regression modeling were performed with a nonparametric technique; the estimated blood loss was replaced with normal scores for this analysis. A stepwise procedure that allowed treatment interaction between covariables and treatment was used for both logistic and linear regression modeling. A value of =.05 was required to enter a variable into a model, except for treatment assignment, which was included in all models. Estimation was performed with least-squares analysis. All analyses were performed with SAS software.¹⁶ A list of candidate variables is provided in the "Appendix."

	Results

Top Abstract Introduction Methods Results Discussion References

 
Baseline and Procedural Characteristics
Baseline and procedural characteristics were similar among the three treatment groups (Table 1). Prestudy aspirin, intravenous heparin, and thrombolytic agents were used by a similar proportion of patients in each treatment group.

View this table: [in this window] [in a new window]   Table 1. Baseline and Procedural Characteristics

Percutaneous coronary revascularization was successfully achieved in >90% of the procedures attempted. The median time spent in the cardiac catheterization laboratory was similar among the three treatment groups. The median total procedural heparin dose was higher in the placebo-treated patients than in the two c7E3-treated groups (P=.001). The proportions of patients who received intravenous heparin beyond 12 hours after the index PTCR were 94%, 88%, and 83% in the placebo, bolus alone, and bolus-plus-infusion groups, respectively (P<.001).

Coagulation and Hematologic Parameters
The median maximum decreases in hemoglobin and hematocrit were 1.8 g/dL and 5.4% in the placebo group, 2.2 g/dL and 6.5% in the bolus group, and 2.3 g/dL and 7.0% in the bolus-plus-infusion group, respectively (P<.001; Table 2).

View this table: [in this window] [in a new window]   Table 2. Changes in Hematologic and Coagulation Parameters

The maximum percent decrements in platelet count occurring 24 hours after study drug infusion were similar in each treatment group. Nadir platelet counts of <100 000 cells/µL and <50 000 cells/µL observed throughout hospitalization tended to occur more often in patients in the bolus-plus-infusion–treated group, but this was statistically insignificant (P=.12). Within the first 24 hours after initiation of study drug infusion, a platelet count <100 000 cells/µL was observed in 7 (1%), 10 (1.4%), and 18 (2.5%) patients in the placebo, bolus, and bolus-plus-infusion groups, respectively (P=.024). Similarly, a greater proportion of patients receiving the bolus-plus-infusion therapy (1.1%) were observed to have a platelet count of <50 000 cells/µL compared with placebo-treated (0.1%) or bolus-treated (0.01%) patients within the first 24 hours after study drug initiation (P=.007).

Bleeding Complications
Major bleeding complications were categorically identified in 222 patients (10.6%): 46 (6.6%), 77 (11.1%), and 99 (14%) in the placebo, bolus, and bolus-plus-infusion groups, respectively (P<.001; Figure). Among patients undergoing CABG, the incidence of major bleeding (P=.81) and estimated units of blood loss (P=.62) were similar among treatment groups. The median times to CABG were 0.8, 1.0, and 2.2 days among patients assigned to the placebo, bolus, and bolus-plus-infusion treatment groups, respectively. Among patients undergoing CABG 24 hours after study drug infusion, the estimated blood loss was 6.9, 7.3, and 8.5 units in the placebo, bolus, and bolus-plus-infusion groups, respectively.

View larger version (21K): [in this window] [in a new window]   Figure 1. Bar graph showing frequency of major and minor bleeding complications among the three treatment groups. Values are given for all bleeding complications (Total) and for bleeding complications not related to coronary artery bypass graft surgery (Non-CABG). Solid bars indicate placebo (n=696); hatched bars, bolus (n=695); and open bars, bolus plus infusion (n=708).

Among the 158 patients with major bleeding events not related to CABG, major bleeding complications were higher in the bolus-plus-infusion group (10.6%) than in the placebo group (3.3%) and were intermediate in the bolus-treated patients (8.6%, Figure). The majority (66.5%) of non–CABG-related major bleeding complications occurred 36 hours from treatment onset (52% placebo, 63% bolus, and 73% bolus-plus-infusion events).

Spontaneous major organ bleeding complications and nonspontaneous bleeding events occurred in 28 (1.3%) and 130 (6.2%) of the patients, respectively, and were significantly more common in patients treated with bolus plus infusion than with placebo (P=.002 and P<.001, respectively, Table 3). Gastrointestinal hemorrhage was the most common type of spontaneous major organ bleeding event and was represented as hematemesis in the majority of cases. The femoral vascular access site was the most frequent site of major bleeding events overall; it was denoted as a bleeding site in 71% (92/130) of the patients who had nonspontaneous hemorrhagic complications. Strokes occurred in 14 patients (0.7%), of which 6 (42.9%) were hemorrhagic. There was no difference in the occurrence of hemorrhagic or nonhemorrhagic strokes among the treatment groups (Table 3). The 95% upper one-sided confidence bounds for hemorrhagic stroke rates were 0.9%, 0.7%, and 1.1% for the placebo, c7E3 Fab bolus, and c7E3 Fab bolus-plus-infusion groups, respectively. Two deaths (0.09%) were attributable to major bleeding from hemorrhagic stroke: one in the placebo group and one in the bolus-plus-infusion group. One additional patient randomized to the bolus-plus-infusion treatment who was not treated with study drug also died of a hemorrhagic stroke.

View this table: [in this window] [in a new window]   Table 3. Location and Type of Events and Transfusion Rates Among Patients With Major Bleeding Complications

Minor bleeding complications not associated with bypass surgery occurred in 291 patients (13.9%) and were more frequent among c7E3 Fab–treated patients (P<.001, Figure). Access site complications accounted for 78% of the total minor bleeding events (placebo, 64.1%; bolus, 83.3%; bolus-plus-infusion, 80.7%). Estimated blood loss for patients with a minor bleeding complication was 3.8, 3.7, and 3.8 units in the placebo, bolus, and bolus-plus-infusion treatment groups, respectively.

Consequences of Major Bleeding Complications
Among patients experiencing major bleeding complications not associated with CABG, the median maximum decrease in hemoglobin (placebo, 6.2 g/dL; bolus, 5.7 g/dL; bolus-plus-infusion, 6.0 g/dL) was similar among the three treatment groups (P=.42). However, blood product transfusions were administered to a greater proportion of c7E3 Fab–treated than placebo-treated patients (placebo, 7.5%; bolus, 14.0%; bolus-plus-infusion, 16.8%; P<.001; Table 3). Packed red blood cells and whole-blood transfusions were administered to 70.3% of all patients with major bleeding complications and to similar percentages of patients in the placebo (60.9%), bolus (70.0%), and bolus-plus-infusion (73.3%) groups (P=.28). The proportion of patients with a major bleeding complication who received >5 units of packed red blood cells was also similar among the three groups (P=.79).

The incidence of dose reduction or discontinuation of study drug related to major non–bypass-related bleeding was low in all groups. There were 21 such patients (3.1%) in the bolus-plus-infusion group, compared with 17 (2.5%) in the bolus group and 2 (0.3%) in the placebo group (P<.001). Surgical intervention was required for a major bleeding complication not related to bypass surgery in 6 placebo patients (26.1% of patients in this group with a major bleeding complication), 12 bolus patients (20.0%), and 5 bolus-plus-infusion patients (6.7%) (P=.007), the majority (17/23, 74%) for femoral vascular repair.

Hospitalization was longer among patients who had a major bleeding complication (median, 7.0 days) than among those without a major bleeding event (median, 3.0 days). However, the length of hospitalization was similar among patients with a major bleeding event, regardless of treatment received (placebo, 7.0 days; bolus, 6.0 days; bolus-plus-infusion, 7.0 days; P=.85).

The incidence of bleeding complications among the first (8.9%), second (6.4%), and third (7.3%) chronological tertiles of patient enrollment was similar.

Factors Associated With Increased Risk of Bleeding Complications
Table 4 shows the results of multivariable logistic regression modeling defining relations of clinical and procedural variables to the occurrence of a major bleeding complication. Age was associated with a major bleeding complication only among patients who received c7E3 Fab, as indicated by the interaction term between age and c7E3 Fab bolus therapy. This interaction term indicates that older patients in the c7E3 Fab–treated groups had a significantly higher rate of major bleeding complications than their placebo-treated counterparts. Other covariables indicating an increased risk of a major bleeding complication included lower weight and presenting with an acute myocardial infarction. The lack of an interaction between weight and trial treatment suggests an increased risk of bleeding complications in low-weight patients regardless of treatment received (Table 4). The odds ratio and P value for c7E3 Fab infusion suggest little difference in the rates of major bleeding complications between the bolus and bolus-plus-infusion groups. After procedural covariables were entered into the model, additional variables associated with higher rates of a major bleeding event included longer PTCR duration, repeat PTCR, and unsuccessful or unattempted coronary revascularization procedures (Table 4).

View this table: [in this window] [in a new window]   Table 4. Clinical and Procedural Variables Associated With Increased Risk of Major Bleeding: Multivariable Logistic Regression Model Results

Table 5 shows the results of the multivariable linear regression model defining the relations of clinical and procedural variables to the maximum drop in hemoglobin (ie, blood loss index). The estimated coefficients represent an indication of the degree of blood loss associated with individual variables (eg, the coefficient -1.11 for acute myocardial infarction indicates an expected 1.11 g/dL greater hemoglobin drop for patients who have this characteristic than for others). No interaction terms were included in the final blood loss index model, suggesting that all risk factors for blood loss were independent of treatment received. Variables associated with an increased amount of blood loss included greater age, female sex, lower weight, higher diastolic blood pressure, higher baseline platelet count, enrollment with acute myocardial infarction or type C lesion morphology, and treatment with c7E3 Fab. Types A and B1 lesions were associated with less blood loss (Table 5). The incremental amount of blood loss associated with the addition of c7E3 Fab infusion to bolus therapy was not statistically significant (P=.428). The addition of procedural variables to the regression model indicated an increased risk of blood loss associated with the same clinical variables but also with longer procedures, intra-aortic balloon pumping, stent use, and CABG, whereas PTCR success was associated with less blood loss (Table 5).

View this table: [in this window] [in a new window]   Table 5. Clinical and Procedural Variables Associated With Increased Blood Loss: Multivariable Linear Regression Model Results

Table 6 lists the odds ratios and 95% confidence intervals for major bleeding complications among patient subgroups. Major bleeding events were more common in women and in older and lower-weight patients. The incremental risk of a major bleeding event among c7E3 Fab bolus-plus-infusion–treated patients was particularly high among women (16.8% versus 4.2%), those who weighed 75 kg (17.8% versus 4.1%), and those who received higher procedural heparin doses per kilogram body weight (11.8% versus 3.3%) compared with the placebo-treated group.

View this table: [in this window] [in a new window]   Table 6. Major Bleeding Among Patient Subgroups

	Discussion

Top Abstract Introduction Methods Results Discussion References

 
Prevalence and Consequences of Blood Loss and Bleeding Complications
Although incremental bleeding risk is not the major determinant of whether a percutaneous coronary revascularization procedure should be performed, knowledge of the underlying risk of bleeding and the incremental risk attributable to a particular therapy should provide part of the basis for decisions about the choice and conduct of the procedure. The study indicates that recent reports of vascular complications in 6% to 29% of patients, resulting in blood product transfusion in 6% to 17% of patients,^{17 18 19} represent a spectrum of risk that could be individualized by knowledge of the incremental risk of specific patient and procedural characteristics. The incremental risk attributable to c7E3 appears to be related primarily to the bolus of the drug and is rarely life-threatening, although c7E3 treatment did lead to increased blood product use. Although we focus only on the details of bleeding risk, the treatment decision must also reflect the previously reported reduction in 30-day ischemic complications and 6-month clinical restenosis rates associated with c7E3 Fab therapy.^{10 11}

In the EPIC trial, a major bleeding complication (not related to bypass surgery) occurred in 7.5% of the total study population. Bleeding complications and estimated units of blood loss were higher in the bolus-plus-infusion c7E3 Fab–treated patients than in the placebo-treated group. However, the majority of hemorrhagic events were transient and clinically well tolerated and did not require surgical intervention. Spontaneous major organ bleeding occurred rarely (1.3%) but was most frequent in the c7E3 Fab bolus-plus-infusion group (2.4%). Spontaneous hematemesis and gross hematuria accounted for the majority of spontaneous major organ hemorrhagic events. The vascular access site was the most common source of major bleeding events overall, occurring in >70% of those experiencing a major bleeding complication in each treatment group. Although the femoral puncture site accounted for the majority of vascular access complications, 16 patients (0.8%) had retroperitoneal hematomas, 12 of whom received bolus-plus-infusion therapy (1.7% of the treatment group).

Stroke occurred in 0.7% of the patients and intracranial hemorrhage in 0.3%, irrespective of treatment assignment. Although the study did not have adequate power to rule out very small differences in risks of hemorrhagic stroke, large differences can be excluded by this study, and the observed results are most consistent with no additional risk from c7E3 Fab therapy. The high-risk characteristics of this population and the careful follow-up evaluations may explain these high rates; conversely, patients with a known history of stroke were excluded. Efforts to reduce these rates will be important in future refinements of percutaneous revascularization procedures and adjunctive therapies.

Blood product transfusions were used in 12.8% of the patients but varied considerably among the treatment groups. The rate of 7.5% among patients receiving placebo is similar to the 6.6% rate among angioplasty patients recently reported.¹⁷ The blood product transfusion rate was approximately twofold higher (14% for bolus; 16.8% for bolus plus infusion) with c7E3 Fab therapy than with placebo. Although transfusion rates were higher in the c7E3 Fab bolus-plus-infusion group than in placebo-treated patients, the extent of hematologic derangement and estimated units of blood loss in patients experiencing a major bleeding event was similar in each treatment group. Differences in procedural techniques (eg, vascular access sheath insertion and removal), dosing of antiplatelet and antithrombotic therapy, complexity of the PTCR procedure, and lack of a specific transfusion protocol may have contributed to the high incidence of major bleeding complications and variable rates of blood product transfusions in the c7E3 Fab–treated patients.

Sparse data are available on estimates of blood loss during PTCR. In two small studies, the absolute decreases from baseline in 12- to 24-hour hematocrit concentrations averaged from 3.8% to 6.0%; this represented a relative decrease of 11% to 15% from the baseline concentration.^{20 21} In the EPIC trial, the median maximum decreases in hemoglobin and hematocrit concentration occurring 36 hours after study drug infusion were 1.8 g/dL and 5.4% in placebo-treated and 2.3 g/dL and 7.0% in bolus-plus-infusion c7E3 Fab–treated patients, respectively. These findings appear comparable to previous studies, despite a higher bleeding risk profile among EPIC patients (Table 1).

Treatment and Procedural Variables Associated With Increased Bleeding Complications
The regression modeling strategy was designed to provide insight into the predicted likelihood of bleeding that could be estimated before the procedure by excluding procedural variables from the initial model. The procedural outcome characteristics were then added to provide an estimate of the effect of conducting the procedure on eventual bleeding outcome. Two measures of bleeding outcome were modeled, since major bleeding, although a discrete and important end point, is not a sensitive measure of blood loss. Indeed, many more factors related to increased risk of blood loss were detected with the models evaluating the adjusted hemoglobin drop. Since no single measure of bleeding outcomes is completely satisfactory, we have found that evaluating several measures provides a worthwhile perspective.

The regression models identified a group of demographic and clinical characteristics that can be used to stratify bleeding risk before a procedure is performed. Based on regression model coefficients, an estimated probability of a major bleeding event can be calculated, given the age and weight of an individual patient, with and without c7E3 Fab therapy. {Complications is derived from estimated coefficients of regression model (Table 4) as a=-1.4199-0.7470+ 0.1633+(0.286xage/10)-(0.245xweight/10) for patients with acute myocardial infarction receiving c7E3 Fab bolus-plus-infusion therapy. For placebo patients not enrolled with acute MI, a=-1.4199-(0.245xweight/10). The probability of an event is exp(a)/[1+exp(a)]. Age is in years, weight in kilograms.} Thus, a 50-year-old patient weighing 90 kg would have a 6% predicted probability of a major bleeding event with c7E3 Fab therapy and a 3% chance without this therapy. In contrast, a 70-year-old patient weighing 70 kg would have an 18% chance of major bleeding with c7E3 Fab and a 4% chance without treatment.

The blood loss index allowed development of a more sensitive index for identifying variables associated with blood loss. Variables associated with the greatest amount of blood loss (>0.5 g/dL drop in hemoglobin) included presentation with acute myocardial infarction (1.04 g/dL decrease), a 35-kg difference in weight (eg, 0.53 g/dL decrease for a 65-kg patient versus a 100-kg patient), greater age, and the presence of complex lesion morphology. Other factors entered into the model were associated with an average drop in hemoglobin of <0.5 g/dL. Blood loss can be expected to increase incrementally and independently with each additional risk factor, regardless of whether c7E3 Fab is given. These same risk factors have been identified in previous studies,^{17 22 23 24 25 26} although the detail provided by this prospective study has not previously been available.

As expected, both the blood loss index and the major bleeding model identified complicated procedures as increasing the risk for transfusion. In the blood loss index model, a procedure duration lasting 100 minutes longer was associated with an average additional hemoglobin decrease of 0.6 g/dL, bypass surgery with a 4.5-g/dL decrease, and stent use with a drop of 1.3 g/dL.

The profound platelet inhibition with c7E3 Fab, when combined with the effects of standard doses of heparin and aspirin, collectively represents a major challenge to hemostatic function, predisposing the patient to PTCR-related hemorrhagic complications. In previous studies, the increasing complexity of PTCR procedures, patient comorbidity, acuteness of patient illness, and concomitant pharmacological therapies (eg, aspirin, intravenous heparin, and fibrinolytic therapy) have been identified as important variables contributing to PTCR-related vascular access site and hemorrhagic complications.^{17 22 23 24 25 26} Within each clinical subgroup, there was an incrementally higher frequency of major bleeding events with the addition of c7E3 Fab bolus and c7E3 Fab bolus-plus-infusion therapy. For example, the frequency of a major bleeding complication among patients weighing 75 kg who received placebo therapy was 4.1% but was approximately threefold higher among patients receiving c7E3 Fab–bolus therapy (10.4%) and fivefold higher (17.8%) when receiving c7E3 Fab bolus-plus-infusion therapy. Similarly, a threefold to fourfold higher frequency of major bleeding complications was observed among those receiving c7E3 Fab bolus and bolus-plus-infusion therapy with higher procedural doses of heparin (120 U/kg, Table 6) than among placebo-treated patients.

The incrementally higher rates of major bleeding complications and blood loss seen in the bolus-plus-infusion group compared with the bolus-only group were neither statistically (Tables 4 and 5) nor clinically (Tables 2 and 3) significant. These observations are important clinically, since the most consistent benefit of c7E3 Fab therapy in high-risk PTCR patients was demonstrable only among those receiving the bolus-plus-infusion regimen.^{10 11} The early intensity of GP IIb/IIIa receptor blockade achieved with either bolus or bolus-plus-infusion dosing of c7E3 Fab²⁷ most likely explains the similar rates of bleeding complications in the two c7E3 Fab regimens.

Conclusions
In the EPIC trial, patients undergoing angioplasty who received adjunctive c7E3 Fab therapy with standard doses of periprocedural aspirin and intravenous heparin had a twofold to threefold higher frequency of major bleeding complications and blood product transfusions than those who received aspirin and heparin alone. Major bleeding complications in the EPIC trial were usually transient and well tolerated. Although major bleeding events were largely attributable to vascular access site complications, they resulted in morbidity, prolonged hospital stays, and increased blood product transfusion regardless of treatment assignment. The sustained GP IIb/IIIa blockade achieved with bolus-plus-infusion c7E3 Fab dosing was not associated with a significant excess of major bleeding complications compared with bolus dosing alone. Major bleeding complications and greater amounts of blood loss occurred more frequently among women, older and lower-weight patients, and those who had complicated or prolonged PTCR. The significant reduction of postprocedural ischemic events observed with c7E3 Fab bolus-plus-infusion therapy strongly warrants consideration of its use among patient subgroups at high risk for post-PTCR ischemic complications. However, clinical risk factors for bleeding complications as described in this report should also be considered in deciding whether to treat a patient with GP IIb/IIIa inhibitors. Ongoing studies of these agents include several efforts to reduce bleeding: reduced heparin dosing, transfusion protocols, premedication with H₂ antagonists, and groin care management guidelines.

View this table: [in this window] [in a new window]   Table 7. Candidate Variables for Regression Models

	Acknowledgments

 
This study was supported by a grant from Centocor, Inc, Malvern, Pa. We wish to express our appreciation to all investigators and clinical coordinators in the EPIC study group. In addition, we express our appreciation to Marilyn J. Utt, Sally Peebles, and Donna Sander for their assistance in preparation of this manuscript; to Trina Stonner, RN, MSN, for her diligent efforts; and to Pat Williams for editorial contributions.

	Footnotes

 
Guest editor for this article was Valentin Fuster, MD, Mount Sinai Medical Center, New York, NY.

¹ A list of EPIC investigators and centers was published in N Engl J Med. 1994;330:956-961.

Candidate Variables for Regression Models
The candidate variables for regression models are given in Table 7.

Received August 25, 1994; revision received December 19, 1994; accepted December 27, 1994.

	References

Top Abstract Introduction Methods Results Discussion References

 

1. Heras M, Chesebro JH, Penny WJ, Bailey KR, Lam JY, Holmes DR, Reeder GS, Badimon L, Fuster V. Importance of adequate heparin dosage in arterial angioplasty in a porcine model. Circulation. 1988;78:640-660. [Abstract/Free Full Text]
   
2. Owen J, Hunter-Laszlo M, Williams JK, Adams M. Thrombin activity induced by balloon angioplasty of the coronary artery in Macaca fascicularis (cynomolgus monkey). Blood Coag Fibrinol. 1990;1:505-507.
   
3. Manolis AS, Melita-Manolis H, Stefanadis C, Toutouzas P. Plasma level changes of fibrinopeptide A after uncomplicated coronary angioplasty. Clin Cardiol. 1993;16:548-552. [Medline] [Order article via Infotrieve]
   
4. Scharf RE, Tomer A, Marzec UM, Teirstein PS, Ruggeri ZM, Harker LA. Activation of platelets in blood perfusion angioplasty–damaged coronary arteries: flow cytometric detection. Arterioscler Thromb. 1992;12:1475-1487. [Abstract/Free Full Text]
   
5. Tschoepe D, Schultheiß HP, Kolarov P, Schwippert B, Dannehl K, Nieuwenhuis HK, Kehrel B, Strauer B, Gries FA. Platelet membrane activation markers are predictive for increased risk of acute ischemic events after PTCA. Circulation. 1993;88:37-42. [Abstract/Free Full Text]
   
6. Schwartz L, Bourassa MG, Lesperance J, Aldridge HE, Kazim F, Salvatori VA, Henderson M, Bonan R, David PR. Aspirin and dipyridamole in the prevention of restenosis after percutaneous transluminal coronary angioplasty. N Engl J Med. 1988;381:1714-1719.
   
7. Barnathan ES, Schwartz S, Taylor L, Laskey WK, Kleaveland JP, Kussmaul WG, Hirshfeld JW. Aspirin and dipyridamole in the prevention of acute coronary thrombosis complicating coronary angioplasty. Circulation. 1987;76:125-134. [Abstract/Free Full Text]
   
8. Ellis SG, Roubin GS, Wilentz J, Douglas JS, King SB. Effect of 18- to 24-hour heparin administration for prevention of restenosis after uncomplicated coronary angioplasty. Am Heart J. 1989;117:777-782. [Medline] [Order article via Infotrieve]
   
9. Walford GD, Midei MM, Aversano TR, Gottlieb SO, Chew PH, Siu CO, Brin KP, Brinker JA. Heparin after PTCA: increased early complications and no clinical benefit. Circulation. 1991;84(suppl II):II-592. Abstract.
   
10. EPIC Investigators. Use of a monoclonal antibody directed against the platelet glycoprotein IIb/IIIa receptor in high-risk coronary angioplasty. N Engl J Med. 1994;330:956-961. [Abstract/Free Full Text]
    
11. Topol EJ, Califf RM, Weisman HF, Ellis SG, Tcheng JE, Worley S, Ivanhoe R, George BS, Fintel D, Weston M, Sigmon K, Anderson KM, Lee KL, Willerson JT. Randomized trial of coronary intervention with antibody against platelet IIb/IIIa integrin for reduction of clinical restenosis: results at six months. Lancet. 1994;343:881-886. [Medline] [Order article via Infotrieve]
    
12. Ellis SG, Roubin GS, King SB III, Douglas JS Jr, Weintraub WS, Thomas RG, Cox WR. Angiographic and clinical predictors of acute closure after native vessel coronary angioplasty. Circulation. 1988;77:372-379. [Abstract/Free Full Text]
    
13. Ryan TJ, Faxon DP, Gunnar RM, Kennedy JW, King SB III, Loop FD, Peterson KL, Reeves TJ, Williams DO, Winters WL Jr, Fisch C, DeSanctis RW, Dodge HT, Reeves TJ, Weinberg SL. Guidelines for percutaneous transluminal coronary angioplasty: a report of the American College of Cardiology/American Heart Association Task Force on Assessment of Diagnostic and Therapeutic Cardiovascular Procedures (Subcommittee on Percutaneous Transluminal Coronary Angioplasty). Circulation. 1988;78:486-502. [Free Full Text]
    
14. Bovill EG, Terrin ML, Stump DC, Berke AD, Frederick M, Collen D, Feit F, Gore JM, Hillis D, Lambrew CT, Leiboff R, Mann KG, Markis JE, Pratt CM, Sharkey SW, Sopko G, Tracy RP, Chesebro JH, for the TIMI Investigators. Hemorrhagic events during therapy with recombinant tissue-type plasminogen activator, heparin, and aspirin for acute myocardial infarction. Ann Intern Med. 1991;115:256-265.
    
15. Landefeld CS, Cook EF, Flatley M, Weisberg M, Goldman L. Identification and preliminary validation of predictors of major bleeding in hospitalized patients starting anticoagulant therapy. Am J Med. 1987;82:703-713. [Medline] [Order article via Infotrieve]
    
16. SAS/STAT User's Guide. Version 6, 4th ed. Cary, NC: SAS Institute, Inc; 1990.
    
17. Grines CL, Glazier S, Bakalyar D, Savas V, Schreiber T, Friedman H, Ellio M, Brodsky M, Sykes C. Predictors of bleeding complications following coronary angioplasty. Circulation. 1991;84(suppl II):II-591. Abstract.
    
18. Hearn JA, King SB III, Douglas JS Jr, Carlin SF, Lembo NJ, Ghazzal ZMG. Clinical and angiographic outcomes after coronary artery stenting for acute or threatened closure after percutaneous transluminal coronary angioplasty. Circulation. 1993;88(pt 1):2086-2096.
    
19. George BS, Voorhees WD III, Roubin GS, Feamot NE, Pinkerton CA, Raizner AE, King SB, Holmes DR, Topol EJ, Kereiakes DJ, Hartzler GO. Multicenter investigation of coronary stenting to treat acute or threatened closure after percutaneous transluminal coronary angioplasty: clinical and angiographic outcomes. J Am Coll Cardiol. 1993;22:135-143. [Abstract]
    
20. Roccario ES, Schweiger MJ, Whitfield SS, Stikley W, Weil T, Urbano A, Porway M. Hematocrit fluctuations after percutaneous transluminal coronary angioplasty. Am J Cardiol. 1991;68:977-978. [Medline] [Order article via Infotrieve]
    
21. Phillips SJ, Spector M, Zeff RH, Skinner JR, Toon RS, Grignon A, Kongtahworn C. Hematocrit changes after uncomplicated percutaneous transluminal coronary angioplasty. Am J Cardiol. 1989;64:940.
    
22. Oweida SW, Roubin GS, Smith RB, Salam AA. Post-catheterization vascular complications associated with percutaneous transluminal coronary angioplasty. J Vasc Surg. 1990;12:310-315. [Medline] [Order article via Infotrieve]
    
23. Kaufman J, Moglia R, Lacy C, Dinerstein C, Moreyra A. Peripheral vascular complications from percutaneous transluminal coronary angioplasty: a comparison with transfemoral cardiac catheterization. Am J Med Sci. 1989;297:222-225.
    
24. Skillman JJ, Kim D, Bain DS. Vascular complications of percutaneous femoral cardiac interventions. Arch Surg. 1988;123:1207-1212. [Abstract]
    
25. Messina LM, Brothers TE, Wakefield TW, Zelenock GB, Lindenauer SM, Greenfield LJ, Jacobs LA, Fellows EP, Grube SV, Steandley JC. Clinical characteristics and surgical management of vascular complications in patients undergoing cardiac catheterization: interventional versus diagnostic procedures. J Vasc Surg. 1991;13:593-600. [Medline] [Order article via Infotrieve]
    
26. Muller DWM, Shamire JK, Ellis SG, Topol EJ. Peripheral vascular complications after conventional and complex percutaneous coronary interventional procedures. Am J Cardiol. 1992;69:63-68. [Medline] [Order article via Infotrieve]
    
27. Coller BS. Platelets and thrombolytic therapy. N Engl J Med. 1990;322:33-42.[Medline] [Order article via Infotrieve]
    

This article has been cited by other articles:

				A. Ali, M. Hashem, H. S. Rosman, L. Moser, A. Rehan, T. Davis, M. Romanelli, T. LaLonde, H. Yamasaki, B. Barbish, et al. Glycoprotein IIb/IIIa Receptor Antagonists and Risk of Bleeding: A Single-Center Experience in 1020 Patients J. Clin. Pharmacol., November 1, 2004; 44(11): 1328 - 1332. [Abstract] [Full Text] [PDF]

				R. J.G. Peters, S. R. Mehta, K. A.A. Fox, F. Zhao, B. S. Lewis, S. L. Kopecky, R. Diaz, P. J. Commerford, V. Valentin, S. Yusuf, et al. Effects of Aspirin Dose When Used Alone or in Combination With Clopidogrel in Patients With Acute Coronary Syndromes: Observations From the Clopidogrel in Unstable angina to prevent Recurrent Events (CURE) Study Circulation, October 7, 2003; 108(14): 1682 - 1687. [Abstract] [Full Text] [PDF]

				J. T. Willerson Myocardial Revascularization with Cardiologic Interventional Devices Card. Surg. Adult, January 1, 2003; 2(2003): 561 - 580. [Full Text]

				R. Chun, B. A. Orser, and M. Madan Platelet Glycoprotein IIb/IIIa Inhibitors: Overview and Implications for the Anesthesiologist Anesth. Analg., October 1, 2002; 95(4): 879 - 888. [Full Text] [PDF]

				L. Y Lee, W. DeBois, K. H Krieger, L. N Girardi, L. Russo, J. McVey, W. Ko, N. K Altorki, R. A Brodman, and O W. Isom The effects of platelet inhibitors on blood use in cardiac surgery Perfusion, January 1, 2002; 17(1): 33 - 37. [Abstract] [PDF]

				H. V. Anderson, J. McNatt, F. J. Clubb, M. Herman, J.-P. Maffrand, F. DeClerck, C. Ahn, L. M. Buja, and J. T. Willerson Platelet Inhibition Reduces Cyclic Flow Variations and Neointimal Proliferation in Normal and Hypercholesterolemic-Atherosclerotic Canine Coronary Arteries Circulation, November 6, 2001; 104(19): 2331 - 2337. [Abstract] [Full Text] [PDF]

				G. J. Despotis, M. S. Avidan, and C. W. Hogue Jr Mechanisms and attenuation of hemostatic activation during extracorporeal circulation Ann. Thorac. Surg., November 1, 2001; 72(5): S1821 - 1831. [Abstract] [Full Text] [PDF]

				S. C. Smith Jr, J. T. Dove, A. K. Jacobs, J. Ward Kennedy, D. Kereiakes, M. J. Kern, R. E. Kuntz, J. J. Popma, H. V. Schaff, D. O. Williams, et al. ACC/AHA guidelines for percutaneous coronary intervention (revision of the 1993 PTCA guidelines): A report of the American College of Cardiology/ American Heart Association Task Force on practice guidelines (Committee to revise the 1993 guidelines for percutaneous transluminal coronary angioplasty) endorsed by the Society for Cardiac Angiography and Interventions J. Am. Coll. Cardiol., June 15, 2001; 37(8): 2239 - 2239. [Full Text] [PDF]

				C. S. Roberts and N. R. Bocanegra Emergent operation for percutaneous coronary rupture after abciximab administration Ann. Thorac. Surg., June 1, 2001; 71(6): 2024 - 2026. [Abstract] [Full Text] [PDF]

				D. P. Chew and D. J. Moliterno A critical appraisal of platelet glycoprotein IIb/IIIa inhibition J. Am. Coll. Cardiol., December 1, 2000; 36(7): 2028 - 2035. [Abstract] [Full Text] [PDF]

				A. M. Lincoff, L. A. LeNarz, G. J. Despotis, P. K. Smith, J. E. Booth, R. E. Raymond, S. K. Sapp, C. F. Cabot, J. E. Tcheng, R. M. Califf, et al. Abciximab and bleeding during coronary surgery: results from the EPILOG and EPISTENT trials Ann. Thorac. Surg., August 1, 2000; 70(2): 516 - 526. [Abstract] [Full Text] [PDF]

				S. C. Silvestry and P. K. Smith Current status of cardiac surgery in the abciximab-treated patient Ann. Thorac. Surg., August 1, 2000; 70(2): S12 - 19. [Abstract] [Full Text] [PDF]

				J. H. Lemmer Jr Clinical experience in coronary bypass surgery for abciximab-treated patients Ann. Thorac. Surg., August 1, 2000; 70(2): S33 - 37. [Abstract] [Full Text] [PDF]

				N. Cauwenberghs, M. Meiring, S. Vauterin, V. van Wyk, S. Lamprecht, J. P. Roodt, L. Novak, J. Harsfalvi, H. Deckmyn, and H. F. Kotze Antithrombotic Effect of Platelet Glycoprotein Ib-Blocking Monoclonal Antibody Fab Fragments in Nonhuman Primates Arterioscler. Thromb. Vasc. Biol., May 1, 2000; 20(5): 1347 - 1353. [Abstract] [Full Text] [PDF]

				D. J. Weiss, M. L. Mirsky, O. A. Evanson, J. Fagliari, D. Mcclenahan, and B. Mccullough Platelet Kinetics in Dogs Treated with a Glycoprotein IIb/IIIa Peptide Antagonist Toxicol Pathol, March 1, 2000; 28(2): 310 - 316. [Abstract] [PDF]

				J. H. Lemmer Jr, M. T. Metzdorff, A. H. Krause Jr, M. A. Martin, J. E. Okies, and J. G. Hill Emergency coronary artery bypass graft surgery in abciximab-treated patients Ann. Thorac. Surg., January 1, 2000; 69(1): 90 - 95. [Abstract] [Full Text] [PDF]