Effect of Pexelizumab, an Anti-C5 Complement Antibody, as Adjunctive Therapy to Fibrinolysis in Acute Myocardial Infarction
The COMPlement inhibition in myocardial infarction treated with thromboLYtics (COMPLY) Trial
Background— Complement activation mediates myocardial damage that occurs during ischemia and reperfusion through multiple pathways. We performed 2 separate, parallel, double-blind, placebo-controlled trials to determine the effects of pexelizumab (a novel C5 complement monoclonal antibody fragment) on infarct size in patients receiving reperfusion therapy: COMPlement inhibition in myocardial infarction treated with thromboLYtics (COMPLY) and COMplement inhibition in Myocardial infarction treated with Angioplasty (COMMA). The COMPLY trial is reported here.
Methods and Results— Overall, 943 patients with acute ST-segment elevation myocardial infarction (MI) (20% with isolated inferior MI) receiving fibrinolysis were randomly assigned <6 hours after symptom onset to placebo, pexelizumab 2.0-mg/kg bolus, or pexelizumab 2.0-mg/kg bolus plus 0.05 mg/kg per h for 20 hours. Infarct size determined by creatine kinase–MB area under the curve was the primary analysis, which included patients who received at least some study drug and fibrinolysis (n=920). The median infarct size did not differ by treatment (placebo, 5230; bolus, 4952; bolus plus infusion, 5557 [ng/mL] · h; bolus versus placebo, P=0.85; bolus plus infusion versus placebo, P=0.81), nor did the 90-day composite incidence of death, new or worsening congestive heart failure, shock, or stroke (placebo, 18.6%; bolus, 18.4%; bolus plus infusion, 19.7%). Pexelizumab inhibited complement for 4 hours with bolus-only dosing and for 20 to 24 hours with bolus-plus-infusion dosing, with no increase in infections.
Conclusions— When used adjunctively with fibrinolysis, pexelizumab blocked complement activity but reduced neither infarct size by creatine kinase–MB assessment nor adverse clinical outcomes.
Received March 13, 2003; revision received June 23, 2003; accepted June 24, 2003.
A series of new fibrinolytic strategies1–4 for acute myocardial infarction (MI) studied in recent trials has failed to improve survival. The continuing and significant morbidity and mortality associated with acute MI underscores the need for new therapies to reduce ischemia and reperfusion injury. Recently, the pathogenic role of inflammation has become appreciated. The complement cascade, activated during acute MI, appears to mediate immune and inflammatory responses in ischemic myocardium.5–7 Complement activation contributes to myocardial necrosis in ischemia/reperfusion through various pathways, including activating leukocytes and endothelial cells; directly damaging cells; and upregulating genes involved in cytokine production, inducible nitric oxide synthase activity, and apoptosis. Accordingly, complement provides a logical therapeutic target.
Pexelizumab is a single-chain fragment of a humanized monoclonal antibody against complement component C5. It specifically binds to C5 with high affinity and prevents cleavage and generation of activated C5a and C5b-9 yet spares the ability to generate C3b, which is critical for opsonizing pathogenic microorganisms and immune complexes. In animals, C5 inhibition has reduced infarct size and apoptosis.8 In phase II clinical studies, C5 inhibition reduced myocardial enzyme release and improved clinical outcomes in patients undergoing cardiopulmonary bypass surgery9 (S. Shernan, MD, unpublished data, 2002).
The Complement And ReDuction of INfarct size after Angioplasty or Lytics (CARDINAL) phase II program tested whether pexelizumab would reduce infarct size, assessed by area under the creatine kinase (CK)-MB release curve at 72 hours, and improve composite clinical outcomes in patients with acute ST-segment elevation MI receiving reperfusion therapy with fibrinolytics—the COMPlement inhibition in myocardial infarction treated with thromboLYtics (COMPLY) trial—or angioplasty—the COMplement inhibition in Myocardial infarction treated with Angioplasty (COMMA) trial. Here we report the results of COMPLY.
Patients were enrolled in 119 centers in 8 countries (Appendix 1). Patients were eligible for recruitment if they were ≥18 years of age, arrived <6 hours after onset of ischemic discomfort that lasted ≥20 minutes, and had ST-segment elevation ≥2 mm in 2 contiguous leads (V1 through V6; I and AVL; or II, III, AVF) or new left bundle-branch block. Exclusion criteria included known hematological dysfunction (hemoglobin <9.5 g/dL, white blood cell count <3000/mm3, neutrophils <1200/mm3, or platelets <100 000/mm3), known or suspected active neisserial infection, known or suspected hereditary complement deficiency, participation in another investigational drug study or exposure to another investigational agent <30 days previously, known or suspected pregnancy or breastfeeding or intention of becoming pregnant during the study, previous renal impairment (serum creatinine >3.0 mg/dL), or evidence of serious active infection. Informed consent was required for participation, and the institutional review board at each site approved the study protocol.
Enrollment proceeded from March 2000 to January 2002. Patients were randomly assigned to treatment by interactive voice randomization, being stratified within each site and by infarct location (isolated inferior MI or noninferior MI according to site investigators). Enrollment of patients with isolated inferior MI was capped at 20% to enroll a high-risk population.
Treatment and Follow-Up
There were 3 study arms: (1) placebo bolus and placebo infusion; (2) 2.0-mg/kg pexelizumab bolus (known to provide complete C5 inhibition for 4 hours)10 and placebo infusion for 20 hours; or (3) 2.0-mg/kg pexelizumab bolus and 0.05-mg/kg per h pexelizumab infusion for 20 hours. The bolus was given before or as soon as possible after the start of fibrinolytic therapy. Infusions began 4 hours after the bolus. The choice of fibrinolytic agent was up to the attending physicians (although alteplase was supplied for use in Brazil and Eastern Europe/Russia), as were other medical therapies.
Patients were followed for in-hospital adverse events and clinical end points (Appendix 2). After discharge, patients were seen at 14, 30, and 90 days. A telephone assessment of vital status occurred at 6 months.
Infarct Size Assessment
CK and CK-MB were measured at enrollment (baseline) and 4, 8, 12, 16, 24, 36, 48, and 72 hours after enrollment. All samples were analyzed at a core laboratory. The primary efficacy variable was infarct size as measured by CK-MB area under the curve (AUC) calculated by the linear-trapezoidal method from data available through 72 hours.11 If baseline or 72-hour values were missing, the corresponding value was set to 0. For missing values of intermediate time points, linear interpolation was used. The primary population for statistical assessment included all patients who received some study drug and some fibrinolytic therapy. For patients who died <72 hours after enrollment, the CK-MB AUC was imputed as the largest total CK-MB AUC value in the study. The median (interquartile range) number of CK-MB samples for patients alive through 72 hours was similar between treatment groups (9 [8, 9] placebo; 9 [9, 9] bolus; and 9 [8, 9] bolus plus infusion). Seven or more samples were collected from 93% of placebo patients, 95% of bolus patients, and 92% of bolus-plus-infusion patients. The number of deaths within 72 hours, and therefore the number of patients with imputed infarct size, was similar between groups (14 placebo, 16 bolus, and 14 bolus-plus-infusion).
Pharmacodynamic and Human Anti-Human Antibody Assays
Blood samples for complement hemolytic activity were drawn on all patients at enrollment and 4, 12, 24, 36, 48, and 72 hours after bolus dosing. Assays for IgG and IgM human anti-human antibodies (HAHA) were performed at a central laboratory on samples collected at enrollment (baseline) and 14, 30, and 90 days after enrollment. Greater than 4-fold increases in HAHA samples from baseline were considered positive.
Twelve-lead electrocardiograms were recorded for all patients at enrollment and 90 minutes and 24 hours after enrollment, at discharge or 6 days after enrollment (whichever occurred first), and at 14 and 30 days after enrollment. Electrocardiograms were evaluated at a core laboratory.
The target sample of 900 patients (300 per treatment arm) was based on 80% power to detect a 20% reduction in CK-MB release. Such a sample also would provide >50% power to detect a one- third reduction in the expected 20% rate of a clinical composite outcome (death, new or worsening heart failure, shock, and disabling stroke at 90 days) in the placebo group, the key secondary end point.
Categorical variables are summarized as percentages, and continuous variables as medians with interquartile ranges. Treatment groups were compared by using stratified log-rank testing or 2-tailed, stratified χ2 tests, and nominal P values (unadjusted for multiplicity) are reported. To preserve the 5% error rate with 2 treatment arms, the primary analyses were to be done sequentially, beginning with placebo versus bolus plus infusion. If a significant difference emerged, the difference between placebo and bolus alone then was to be calculated using a closed-test procedure. Kaplan-Meier survival curves were compared by using survival analysis methods.
Subgroups were specified based on geographic region, MI location, age, sex, diabetes status, previous MI, Killip class, fibrinolytic agent used, and race. Logistic regression analysis was performed on the clinical composite to evaluate the effects of potential differences in baseline covariates, including geographic region, MI location, age, sex, diabetes status, previous MI, heart rate, blood pressure, Killip class, and fibrinolytic type.
In all, 943 patients were enrolled (Figure 1). Baseline characteristics were similar between treatment groups (Table 1). Study drug was started a median 10 minutes after fibrinolytic therapy. Concomitant medication use was comparable between the groups (Table 2).
As anticipated, the pexelizumab bolus and bolus plus infusion almost completely inhibited complement for 4 hours and 20 to 24 hours, respectively (Figure 2). Positive responses in the IgG and IgM HAHA were rare (0 of 242 and 1 of 242 placebo, 1 of 223 and 3 of 223 bolus, and 1 of 238 and 9 of 238 bolus-plus-infusion patients, respectively, at 14 days, and 2 of 232 and 2 of 232 placebo, 0 of 223 and 0 of 223 bolus, and 1 of 240 and 0 of 240 bolus-plus-infusion patients at 90 days).
This primary end point did not differ significantly between the groups: placebo, 5230 (2492 to 9054); bolus-only, 4952 (2033 to 8522); bolus-plus-infusion, 5557 (2445 to 9333) [ng/mL] · h; P=0.85 for bolus-only versus placebo; P=0.81 for bolus-plus-infusion versus placebo (Figure 3). The 95% CIs corresponded to a small likelihood of a >8% decrease in median infarct size. Median peak CK-MB levels also did not differ markedly between groups: placebo, 242 (101 to 459); bolus-only, 214 (78 to 412); bolus-plus-infusion, 244 (95 to 433) ng/mL.
Clinical End Points
The incidence of the 90-day composite clinical end point was similar between treatment groups (Table 3). Most events occurred early in each group. Somewhat fewer disabling strokes had occurred by 90 days in pexelizumab-treated patients. The proportion of patients with complete (>70%) ST-segment resolution, results of QRS score analyses, and rates of other clinical events were similar between the treatment groups (Table 4).
In subgroup analysis, North Americans (n=434) who received bolus-plus-infusion dosing had a nonsignificant reduction in the 90-day composite clinical outcome compared with the placebo group (relative risk [RR], 0.82; 95% CI, 0.49 to 1.38) (Figure 4). Multivariable regression analysis showed no difference in the adjusted and unadjusted 90-day clinical event rates (unadjusted P=0.836; adjusted P=0.51).
Pexelizumab was well tolerated, causing no increase in adverse events. Infectious complications were infrequent. Serious infection was reported in 6 placebo, 4 bolus, and 2 bolus-plus-infusion patients and was most often pneumonia. Left ventricular mechanical complications also were rare (myocardial rupture occurred in 6 placebo patients, 3 bolus patients, and 5 bolus-plus-infusion patients).
In this trial, complement inhibition with pexelizumab as an adjunct to fibrinolysis had no measurable effect on the primary end point, infarct size by CK-MB AUC through 72 hours, or on clinical end points through 90 days.
Inflammation, measured by markers such as C-reactive protein12,13 or leukocytosis,14 relates to poor outcomes in acute MI. Although inhibiting inflammation with corticoster- oids15,16 or nonsteroidal inflammatory agents16–18 has not been shown to improve outcome and may even impair infarct healing, more targeted antiinflammatory therapies have reduced infarct size in animal models of ischemia and reperfusion. Recently, rhuMAb CD 18, an inhibitor of leukocyte adhesion, failed to improve coronary perfusion measured by Thrombolysis In Myocardial Infarction (TIMI) frame count or reduce infarct size in patients with acute MI.19 One possibility regarding the treatment failure in humans is that inhibition of one element of the inflammatory cascade is ineffective.
The complement system is a promising therapeutic target because it activates multiple inflammatory processes. Its terminal membrane attack complex is present in animal and human infarcted myocardial tissue.20 Complement inhibition has reduced infarct size in several animals, including the baboon,20 pig,21 and rat.8 Complement inhibition with C5 antibody also has decreased neutrophil infiltration and apoptosis in rats.8 In humans, complement inhibition with a C1-inhibitor for 48 hours was well tolerated and associated with lower than expected troponin and CK-MB measurements in 22 patients given streptokinase for acute MI.22 In a phase II study of pexelizumab in cardiopulmonary bypass surgery, treatment with the same doses used in this study resulted in similar inhibition of complement activity. In a secondary analysis, death and marked CK-MB elevation were significantly reduced with bolus-plus-infusion dosing9 (S. Shernan, MD, unpublished data, 2002).
Multiple strategies have been used to evaluate myocardial infarct size, and each has its advantages and limitations. Infarct size measured by serial CK-MB assessment correlates with left ventricular failure,23,24 presence of Q waves,25 and mortality,24–26 although these findings are debated. Nevertheless, cardiac-enzyme release has been used to show a reduction in infarct size by streptokinase treatment versus placebo,27 alteplase versus streptokinase,28 and angioplasty versus streptokinase.29 Such surrogates are useful for establishing proof of concept for new therapies but have inherent limitations in predicting treatment effect on clinical outcomes. Modulating inflammatory pathways during MI and reperfusion could enhance outcomes in ways not evident from early infarct size measurements.
In this trial, complete enzyme data were available for 90% of patients, and given the distribution of infarct size measurements, we are confident that infarct size was not significantly affected. Moreover, the CK-MB data were consistent with electrocardiographic data, which showed no difference by treatment in either myocardial reperfusion by early ST-segment resolution or infarct size by QRS score.
Pexelizumab was associated with no improvement in the clinical composite of death, shock, heart failure, or disabling stroke. However, the trial was not powered to evaluate clinical outcomes, and there was a clinically relevant reduction in the clinical composite at 90 days with pexelizumab bolus plus infusion compared with placebo (RR, 1.06; 95% CI, 0.77 to 1.47). Active treatment also was associated with a nonsignificant reduction in disabling stroke, compatible with a potential neuroprotective effect noted after bypass surgery in patients given pexelizumab.9
The trial enrolled a high-risk patient population by limiting patients with isolated inferior MI to 20% overall, and all patients had ≥2 mm ST-segment elevation in 2 contiguous leads. Mortality at 30 days was 8.7% overall, higher than that seen in several contemporary acute MI trials (5.4% to 6.6%).2–4 In small trials, baseline imbalances can occur that might affect interpretation of clinical effects, and in COMPLY, the bolus-plus-infusion group did have greater baseline risk, including more advanced age, higher Killip class, and more prior infarction. Even in multivariable modeling, however, the composite event rate was similar between treatment groups.
There was no statistically significant interaction of major subgroup with treatment effect of pexelizumab. In North American patients, however, the point estimates for both the clinical composite and mortality at 90 days trended toward improved outcome with pexelizumab bolus plus infusion relative to placebo.
To prevent delays in initiating reperfusion therapy, study-drug administration was allowed to begin after fibrinolysis, which could have reduced pexelizumab’s benefits. This effect likely would be attenuated, however, by the lag between the start of pharmacological therapy and actual reperfusion. Pexelizumab treatment also might be more effective with longer treatment. Finally, although complement activity assessed by hemolytic activity was successfully suppressed, additional analyses of pexelizumab’s effects on other inflammatory mediators regulated by C5a and C5b-9 might prove valuable.
In conclusion, pexelizumab had no measurable effect on infarct size assessed by CK-MB release or on clinical outcomes when used as an adjunct to fibrinolysis. This novel therapy was well tolerated and was not associated with increased risk of infection or myocardial rupture.
C.B. Granger, K.W. Mahaffey, P.W. Armstrong, J.S. Hochman, C.F. Mojcik, J.C. Nicolau, W. Ruzyllo, P. Theroux, T.G. Todaro, W.D. Weaver.
Data Safety and Monitoring Committee
T. Ryan, C.P. Cannon, K. Bailey, P.S. Berger, N.S. Kleiman.
T. Filloon, V. Hasselblad, G. Tasissa. Project Managers: K.J. Malloy, A. Tummon Kamphuis, E. Fraulo, B. Maeland, E. James, M. Adam, D. Sydlowski, A. Wynne.
Numbers of patients enrolled are shown in parentheses.
Brazil (78): L. Maia, J. Saraiva, A. Carvalho, G. Greque, J. Nicolau, P. Nogueira, V. Golin, J. Filho, D. de Albuquerque, A. do Amaral Baruzzi. Canada (217): A. Hess, C. Benke, G. Gosselin, M. David, Y. Pesant, C. Proulx, M. Senaratne, K. Heck, J. Kornder, M. Perry, H. Comtois, R. Major, S. Dhingra, D. Anderson, V. King, V. Sawchuk, L. Kuspira, P. Chagnon, L. Bourgeois, L. Gauthier, D.-K. Nguyen, D. Racicot, P. Theroux, N. Gendron, C. MacCallum, M. Constantine, G. Pruneau, F. Ouimet, A. Weeks, S. Stoger, V. Gebhardt, J. Taylor, P. Greenwood, A. Prosser, J. Heath, L. Saffarek, P. Polasek, V. Stedham, M. Richmond, S. Essmail, P. Auger, N. Bilodeau, F. Dumont, S.N. Ali, C. DeGroot, L. Robichaud, S. Carrier, J. Piché, L. Desjardins, L. St-Pierre, R. Brossoit, C. Robert, D. Gossard, L. Boutin, F. Sandrin, S. Mitges, C. Wells, K. Newton, J. Smith, M. Weigel, P. Chisholm, D. Kincade, L. Kembel, M. Kincade, F. Ervin, R. Ackerman, M. Turabian, L. Hunter, D. Winter, J. Milton, J. Drake, G. Kumar, D. Zippel, N.K. Sharma, L. Bergen. Estonia (26): Ü. Soopold, J. Voitk. Hungary (58): S. Timar, B. Oze, Z. Laszlo, Z. Pecsvarady. Lithuania (22): A. Laucevicius, J. Brazdzionyte, A. Kirkutis. Poland (232): K. Jaworska, J. Kopaczewski, M. Kurowski, W. Pluta, A. Malinski, D. Wojciechowski, P. Kolodziej, W. Banasiak, M. Krzciuk, A. Hoffman. Russia (73): M. Ruda, M. Glezer, G. Arutyunov. United States (237): G.R. McKendall, M. Grogan, M. Nallasivan, R. Arguelles, P. Caples, K. Loxterkamp, G. LaMar, S. Larsen, B. Lupi, D. Weinberg, D. Wilson, A. Rees, D. Alexander, V. Bethala, S. Bennett, K. Cannon, K. Mahaffey, C. Martz, P. Gottleib, J. Painter, K. DeVilbiss, C. Overstreet, K.A. Klancke, M. Bayer, W.J. Rogers, T. Holmes, T. Morgan, E. Rivera, K. Pulliam, J.J. Chahin, N. Bruno, J. Foley, P. Provost, A. Rashkow, J. Anderson, N. El Sanadi, L. Siebert, A. Niederman, T. Kellerman, N. Dhruva, L. Patten, S. Promisloff, D. Collette, P. Gordon, N. Wright, L. Garza, S. Brock, C. Lundergan, A.M. Cangialosi, T. Hilton, S. Duncan, M. Imburgia, K. Hulsmeyer, P. Tucker, M. Leesar, B. Van Hoose, M. Solovay, J. Miller, J. Stachler, J. Greenwood, K.H. Sheikh, T. Hengerer, R. Vicari, M. Howard, J.C. Field, D. Dvorak, G. Schaer, J. Bax, J. Martinez, B. Jubelin, E. Hockstad, B. Lee, M. Bernat, P. Johnson, R. Frazier, N. Slemmons, M. Sharma, R. Cooper, O. Rosales, C. Underwood, C. Carter, K. Akosah, L. Storlie, R. Reeves, D. Fowler, J. Hargrove, M. Lynch, C. Foley, J. Saucedo, S. Brock, K. Jutzy, V. Bishop, R. Prashad, C. McDonough, E. Rogers, B. Gay, D. Banish, V. Nicely, C. Reams, C.R. Norman, J. Boe, R. Dunne, D. Waselewsky, M. Hudson, J. Cote, S.D. Stenstrom, C. Chamernick, R. Bach, K. Luepke, R. Montgomery, A. Ontko, J. Daniels, G. Chaikin, D. Chang, A. Todd, C. Schaeffer, J. Mandak.
Clinical End Point Definitions
Congestive Heart Failure (occurring ≥24 hours after enrollment, including rehospitalization for heart failure)
Physician decision to treat with diuretic, intravenous inotropic agent, or intravenous vasodilator and one or both of the following: (1) pulmonary edema or vascular congestion on chest x-ray believed to be of cardiac cause, or (2) any 2 or more of the following: rales greater than one-third up the lung fields believed to be from heart failure; pulmonary capillary wedge pressure >18 mm Hg; dyspnea with recorded Po2 <80 mm Hg on room air or o2 saturation <90% on room air, without significant lung disease; increasing dyspnea on exertion (for rehospitalization); nocturnal dyspnea (for rehospitalization); or worsening peripheral edema (for rehospitalization).
Systolic blood pressure <90 mm Hg for ≥1 hour, unresponsive to fluid replacement alone, considered secondary to cardiac dysfunction, and associated with cool, clammy skin; oliguria; altered sensorium; or a cardiac index ≤2.2 L/min per m2. If blood pressure increased to ≥90 mm Hg with positive inotropic or vasopressor agents alone or with mechanical support in <1 hour, this was classified as cardiogenic shock.
New focal neurological deficit lasting >24 hours with moderate or severe deficit or resulting in death and classified as a stroke by a physician. Patient functional status (deficit), as follows, was determined at discharge or 30 days (whichever came first): none (no sequelae), minor (functional status unchanged), moderate (significant limitation of activity/capabilities), or severe (unable to live independently or work). Disabling stroke resulted in moderate or severe impairment.
Funding for this study was provided by Procter & Gamble Pharmaceuticals and Alexion Pharmaceuticals, Inc. We acknowledge Dr Kevin Malloy for his important roles in trial design, project leadership, data interpretation, and manuscript review; Dr Vic Hasselblad for help with trial design and analysis; and Dr Jennifer King and Anthony Doll for help with manuscript preparation.
Drs Filloon and Todaro are employees of Procter & Gamble Pharmaceuticals. Dr Mojcik is an employee of Alexion Pharmaceuticals, Inc. Drs Mahaffey, Granger, Nicolau, Ruzyllo, Weaver, Theroux, Hochman, and Armstrong have received consultation fees and/or research grants from Procter & Gamble Pharmaceuticals and Alexion Pharmaceuticals, Inc.
Yasuda M, Takeuchi K, Hiruma M, et al. The complement system in ischemic heart disease. Circulation. 1990; 81: 156–163.
Vakeva AP, Agah A, Rollins SA, et al. Myocardial infarction and ap- optosis after myocardial ischemia and reperfusion: role of the terminal complement components and inhibition by anti-C5 therapy. Circulation. 1998; 97: 2259–2267.
Fitch JC, Rollins S, Matis L, et al. Pharmacology and biological efficacy of a recombinant, humanized, single-chain antibody C5 complement inhibitor in patients undergoing coronary artery bypass graft surgery with cardiopulmonary bypass. Circulation. 1999; 100: 2499–2506.
Anzai T, Yoshikawa T, Shiraki H, et al. C-reactive protein as a predictor of infarct expansion and cardiac rupture after a first Q-wave acute myocardial infarction. Circulation. 1997; 96: 778–784.
Barron HV, Cannon CP, Murphy SA, et al. Association between white blood cell count, epicardial blood flow, myocardial perfusion, and clinical outcomes in the setting of acute myocardial infarction: a Thrombolysis In Myocardial Infarction 10 substudy. Circulation. 2000; 102: 2329–2334.
Hammerman H, Kloner RA, Hale S, et al. Dose-dependent effects of short-term methylprednisolone on myocardial infarct extent, scar formation, and ventricular function. Circulation. 1983; 68: 446–452.
Hammerman H, Kloner RA, Schoen FJ, et al. Indomethacin-induced scar thinning after experimental myocardial infarction. Circulation. 1983; 67: 1290–1295.
Baran KW, Nguyen M, McKendall GR, et al. Double-blind, randomized trial of an anti-CD18 antibody in conjunction with recombinant tissue plasminogen activator for acute myocardial infarction. Circulation. 2001; 104: 2778–2783.
Stahl GL, Amsterdam EA, Symons JD, et al. Role of thromboxane A2 in the cardiovascular response to intracoronary C5a. Circ Res. 1990; 66: 1103–1111.
de Zwaan C, Kleine AH, Diris JC, et al. Continuous 48-h C1-inhibitor treatment, following reperfusion therapy, in patients with acute myocardial infarction. Eur Heart J. 2002; 23: 1670–1677.
Sobel BE, Bresnahan GF, Shell WE, et al. Estimation of infarct size in man and its relation to prognosis. Circulation. 1972; 46: 640–648.
Bleifeld W, Mathey D, Hanrath P, et al. Infarct size estimated from serial serum creatine phosphokinase in relation to left ventricular hemodynamics. Circulation. 1977; 55: 303–311.
Baardman T, Hermens WT, Lenderink T, et al. Differential effects of tissue plasminogen activator and streptokinase on infarct size and on rate of enzyme release: influence of early infarct related artery patency. Eur Heart J. 1996; 17: 237–246.
de Boer MJ, Suryapranata H, Hoorntje JC, et al. Limitation of infarct size and preservation of left ventricular function after primary coronary angioplasty compared with intravenous streptokinase in acute myocardial infarction. Circulation. 1994; 90: 753–761.