acadesine: a new drug that may improve myocardial … · 2016. 12. 12. · however, acadesine re-...

ACADESINE: A NEW DRUG THAT MAY IMPROVE MYOCARDIAL PROTECTION IN CORONARY ARTERY BYPASS GRAFTING Results of the first international multicenter study

The effect of acadesine, an adenosine-regulating agent, on the incidence of myocardial infarction, all adverse cardiovascular outcomes (myocardial infarction, cardiac death, left ventricular dysfunction, life-threatening arrhythmia, or cerebrovascular accident) and mortality was assessed in 821 patients undergoing coronary artery bypass grafting. Patients were prospectively stratified to a high-risk group (age >70 years, unstable angina, previous coronary bypass, unsuccessful angioplasty, or ejection fraction <30%) or a non-high-risk group. They were randomized in a double-blind manner to placebo (n = 418) or acadesine (n = 403) by intravenous infusion over 7 hours (0.1 mg/kg per minute) and in the cardioplegic solution (placebo or acadesine; 5 /xg/ml). Acadesine did not significantly affect the incidence of myocardial infarction in the overall study population, but it significantly reduced the incidence of Q-wave myocardial infarction in high-risk patients (placebo, 19.7%; acadesine, 10.0%; p = 0.032). The incidences of all adverse cardiovascular outcomes (placebo, 19.4%; acadesine, 18.4%) and overall mortality (placebo, 3.4%; acadesine, 2.7%) were similar between the two treatment groups. However, acadesine reduced the incidence of cardiac related events that contributed to deaths occurring during the first 3 postoperative days so that the incidence of death in this period was lower (placebo, 1.9%; acadesine, 0.2%; p = 0.038). No adverse events were related to acadesine treatment. Although overall there were no statistically significant between-group differences for the primary study end points, a secondary analysis in a prospectively defined high-risk subgroup suggests that acadesine may be beneficial in some patients. (J THORAC CARDIOVASC SURG 1995;110:1096-106)

Philippe Menasch6, MD, PhD, Paris, France, W. R. Eric Jamieson, MD, Vancouver, B.C., Canada, Willem Flameng, MD, PhD, Leuven, Belgium, and Michael K. Davies, MD, Birmingham, United Kingdom, on behalf of the Multinational Acadesine Study Group*

M orbidity and mortality after coronary artery bypass grafting (CABG) have increased over

the past decade, despite improvements in surgical

From the Service de Chirurgie Cardio-Vasculaire, H6pital Lari- boisi6re, Paris, France, the Department of Surgery, University of British Columbia, Vancouver, B.C., Canada, the Depart- ment of Cardiac Surgery, University Hospital Gasthuisberg, Leuven, Belgium, and the Division of Medicine, Selly Oak Hospital, Birmingham, United Kingdom.

Supported by a grant from Gensia, Inc., San Diego, Calif. Presented in part, in abstract form, at the 1993 Annual Meeting

of the American College of Cardiology, Anaheim, Calif., and the Fifteenth Congress of the European Society of Cardiol- ogy, 1993, Nice, France.

Received for publication April 20, 1994. Accepted for publication Jan. 9, 1995. Address for reprints: Philippe Menasch6, MD, PhD, Service de

Chirurgie Cardio-Vasculaire, H6pital Lariboisi6re, 2, Rue Ambroise-Par6, 75475 Paris, France.

1 0 9 6

techniques and perioperative management . Several studies have demonstrated that older patients with more extensive coronary disease, poorer left ventricular function, and more comorbid diseases are now undergoing CABG and that these factors have a negative impact on morbidity and morta l i ty) -8 In recent studies, representative of the current patient population undergoing CABG and surgical prac- tices being used, the mortality rate ranged from 3% to 8%. 1-3'9' 10 Cardiac events are still the most common causes of death after CABG, accounting for approximately 55% of the deaths, n Of these cardiac events, perioperative myocardial infarction (MI) is a significant complication of CABG, and its

*The investigators and institutions participating in the Multina- tional Acadesine Study are listed in the appendix.

Copyright © 1995 by Mosby-Year Book, Inc. 0022-5223/95 $5.00 + 0 12/1/63178

brought to you by COREView metadata, citation and similar papers at core.ac.uk

provided by Elsevier - Publisher Connector

https://core.ac.uk/display/82468692?utm_source=pdf&utm_medium=banner&utm_campaign=pdf-decoration-v1

The Journal of Thoracic and Cardiovascular Surgery Volume 110, Number 4, Part 1

MenaschO et aL 1 0 9 7

occurrence has been shown to have an adverse effect on postoperative mortality rates and long-term survival, a2-17 The prevalence of perioperative MI after CABG has been reported in the literature to range f rom 4% to 26%, depending on the criteria used for diagnosis.Z, 15, 17-3o The unequivocal recognition that the risk of morbidity and mortality associated with CABG can still be reduced has therefore led to the development of new strategies designed to preserve the myocardium by reducing ischemia and reperfusion injury. 31"3z Although some approaches have focused on the refinement of cardioplegic formula- tions, the tight control of reperfusate composition, or the search for the optimal route and technique of delivering cardioplegic solutions, an alternative ap- proach has consisted of supplying the myocardium with substrates or agents designed to reduce the extent of ischemia-reperfusion injury. In the latter case, however, the expected achievement of a therapeutic benefit has consistently relied on an exogenous supply of the purported cardioprotective agents. More recently, efforts have been focused on managing endogenous systems to lessen myocardial ischemia and necrosis. Endogenous adenosine is an important mediator of myocardial preservation because activation of adenosine receptors improves myocardial survival after ischemic insults (i.e., ischemic preconditioning). 33

Acadesine (5-amino-l-[/3-D-ribofuranosyl]imida- zole-4-carboxamide) belongs to a new class of agents called adenosine-regulating agents. These agents are believed to act by locally increasing the availability of adenosine only in tissues depleted of adenosine triphosphate (i.e., ischemia). 34-36 Several animal studies have now documented the cardioprotective effects of acadesine, including under conditions of global ischemic arrest, as occurs during CABG in human beings. 37' 38 In an initial small clinical trial in 116 patients undergoing elective CABG, acadesine given intravenously and in cardioplegic solutions was shown to reduce the incidence of perioperative MI by 64% compared with placebo treatment. 39 Although the small sample size probably prevented attainment of statistical significance in this early study, the results were suggestive and provided the rationale for this larger, multinational randomized trial.

Methods

General study design. A double-blind trial was conducted in 821 patients with coronary artery disease. The patients were stratified by risk group and randomized in this placebo-controlled, parallel-group study. All were

scheduled for elective, urgent, or emergency CABG and were enrolled at 27 medical centers in Europe and Canada (appendix). The protocol was approved by the ethics committee at each center. In accordance with the principles of the Helsinki Declaration, written informed consent was obtained from each patient. Patients received either placebo or acadesine administered intravenously for a total of 7 hours, beginning 15 minutes before the induction of anesthesia. The duration of infusion included the intraoperative (prebypass and postbypass) period and the immediate postoperative period in the intensive care unit for virtually all patients. The cardioplegic solution used for myocardial protection during cardiopulmonary bypass contained acadesine (final concentration of 5 /xg/ ml) or placebo, as appropriate to match the randomization for intravenous drug administration. An independent safety and data monitoring panel was responsible for decisions about the safe conduct and continuation of the trial.

Patient selection. Patients were enrolled in the study between June 1991 and June 1992. Patients of either sex were eligible for recruitment. To be included in the study, patients were required to have coronary artery disease as assessed by coronary angiography, to have left main stenosis of at least 50% or stenoses of at least 70% of two or more major coronary arteries, to be at least 35 years of age, and to be in hemodynamically stable condition at the time of the operation. Excluded from the study were the following: patients having valve replacement or hemodynamically significant valvular disease, patients with suspected or definite recent or_ evolving MI (within 7 days before the operation as evidenced by creatine kinase myocardial band [CK-MB] measures above the screening laboratory's reference range or changes in Q-wave morphology), or patients with a history of uric acid nephropathy or hyperuricemia (serum uric acid value 20% above the institution's laboratory reference range), renal insufficiency (creat- inine concentration greater than two times the upper limit of the testing laboratory's reference range at the time of screening), or hepatic dysfunction (alanine aminotransferase or aspartate aminotransferase concentrations greater than three times the upper limit of the testing laboratory's reference range at screening).

Stratification, randomization, and treatment. Before randomization, patients were prospectively stratified into one of two groups: (1) high risk and (2) all other patients, termed non-high risk. The criteria defining high risk were age greater than 70 years, previous CABG, acute failure of percutaneous transluminal coronary angioplasty (without an evolving M1), unstable angina, and left ventricular dysfunction (ejection fraction <30%). Patients were ran- domly allocated within each center to treatment with placebo (Sterile Water for Injections, British Pharmaco- poeia) or acadesine (0.1 mg/kg per minute) according to a previously prepared computer-generated randomization code for each risk group.

Adenosine, theophylline, aminophylline, pentoxifylline, and dipyridamole were not to be given for the 48 hours before the operation and during the first 48 hours after the operation because their presence might alter any

1 0 9 8 Menasch~ et al. The Journal of Thoracic and

Cardiovascular Surgery October 1995

effects of acadesine. Intravenous nifedipine or other intravenous calcium-channel antagonists were permitted, except as antiischemic prophylaxis. All other cardiovascular medications were continued, as indicated, until the operation. Anesthesia induction was limited to intravenous fentanyl, alfentanyl, sufentanyl, midazolam, or propofol, according to local practice. Fluorinated inhala- tional agents were prohibited except for the use of isoflu- rane at one center. Operations were performed by means of standard techniques for myocardial revascularization. The type and method of administration of cardioplegic solutions were not controlled; however, cardioplegic solutions containing drugs other than the study drug (e.g., lidocaine) were not to be used. The use of warm blood cardioplegia was generally discouraged, although a subset of patients was treated with warm cardioplegia or with single warm blood terminal infusions ("hot shots"). One center used the surgical technique of intermittent aortic crossclamping without the use of any cardioplegic solutions.

Outcome measurements and follow-up. The outcome measurements in this study were the incidence of fatal and nonfatal MI and the incidence of all adverse cardiovascular outcomes (MI, cardiac death, severe left ventricular dysfunction, life-threatening arrhythmia, or cerebrovascular accident). Outcomes were assessed for their occurrence through hospital discharge. Follow-up for death and any other major adverse cardiovascular event was continued for up to 12 months after the operation at 22 of the 27 participating centers.

Perioperative MI was diagnosed according to the following criteria: (1) the presence of a new Q wave on postoperative twelve-lead electrocardiograms (ECGs) using Minnesota code criteria 15 with clinical overreading; (2) elevation of serum CK-MB concentration to 100 ng/ml or more at any time after the operation, to 70 ng/ml or more at any time after 12 hours after the operation, or to 12 ng/ml or more more than 24 hours after the operation.ZV'4°'41; or (3) MI on autopsy. Twelve-lead ECGs were obtained at screening (within 7 days before the operation), on arrival in the intensive care unit, on postoperative days 1, 2, 3, and 4, and at hospital discharge. ECGs were transferred to a central site where they were coded and evaluated by a panel of blinded reviewers unaware of any other study data, comprised of a senior cardiologist and several coders. The senior cardiologist reviewed all sets of ECGs using standard ECG interpretive techniques to assess whether or not an MI had occurred. Two separate individuals independently assigned Minnesota codes to each ECG and additionally reviewed each set using standard ECG interpretive techniques to assess whether a new Q-wave MI was present. ST-T wave changes unsupported by localized depolariza- tion abnormalities (e.g., new Q wave, deeper or wider Q wave) were not considered sufficient for the diagnosis of a perioperative MI. The panel, consisting of the senior cardiologist and all coders, reviewed each coded ECG set and agreed on the presence or absence of an MI for each patient. CK-MB concentrations were measured in serum samples obtained at screening and at 1, 4, 8, 12, 16, 20, 24, 28, 32, 36, 42, 48, and 60 hours after removal of the aortic crossclamp. Processed serum samples were stored at

- 2 0 ° C at each site and then sent to the central laboratory. The samples were stored at - 2 0 ° C at the central laboratory until blinded analysis was performed by means of an immunoenzymetric assay kit (Tandem-E CKMB II, Hybritech [USA], San Diego, Calif.). Autopsy diagnosis of MI was made by the pathologist at each institution.

Because of the complexity of diagnosing MI after CABG, MI may be defined by one or more of the criteria for diagnosis described earlier. Results are presented with less specific criteria (presence of a Q wave or elevation of CK-MB concentrations, or autopsy MI) used for the power calculation for this study, more specific criteria (Q wave and CK-MB, or autopsy MI), and by ECG and autopsy criteria alone (Q wave or autopsy MI). The latter secondary analysis was performed because the CK-MB data set was not considered completely reliable in this study; the CK-MB concentrations were lower than expected, which is inconsistent with literature reports in the setting of CABG and the lack of CK-MB elevations in several patients with MIs proved at autopsy. Reasons for the questionable reliability of the CK-MB data set are discussed later.

For patients who died, the cause of death was classified as to whether it was a primary cardiac death. Classifica- tions were determined by an independent panel of three cardiologists not associated with the clinical study, who were blinded to the patient 's treatment assignment. Death occurring within the first 3 days after a CABG operation is considered acute11; therefore the effect of acadesine on mortality during this period was also assessed.

Severe left ventricular dysfunction was considered present if cardiogenic shock (defined as a cardiac index <1.5 L/min per square meter and pulmonary capillary wedge pressure >20 cm for >1 hour) occurred or if an intraaortic balloon pump or left ventricular assist device for low output syndrome was required.

Life-threatening arrhythmia was considered present if either of the following two criteria were met: (1) ventricular arrhythmia necessitating cardioversion, other than during bypass; and (2) conduction defect necessitating insertion of a permanent pacemaker. Cardioversion for atrial arrhythmias was not included in this definition of an adverse cardiovascular outcome.

A cerebrovascular accident was considered present if signs or symptoms of significant neurologic deficit per- sisted for more than 24 hours. Focal neurologic deficits associated with local nerve injury (e.g., local neuropathies produced by positioning during the operation) were not considered cerebrovascular accidents. Patients with non- focal deficits were classified as having a cerebrovascular accident only when the diagnosis was confirmed by neurologic examination or when results from computed to- mographic scan or magnetic resonance imaging were consistent with a new cerebral infarct or hemorrhage.

Statistical analysis. The sample size calculation of approximately 400 patients per treatment group was made by the Lachin method, assuming two treatment groups and two outcomes (MI and no MI), with the underlying model being a 2 by 2 contingency table on which a X 2 test would be performed. 42 The sample size was calculated to give a 90% chance of detecting a 33% reduction in the incidence of MI with a level of significance at 5%. The


MenaschO et al. 1 0 9 9

placebo rate for the incidence of MI was set at 30%. A total of 53 strata, defined by combinations of 27 centers and two risk groups (one center had no high-risk patients), were used. Nine patients were assigned at study sites into the incorrect risk stratum. These patients were reclassified into the correct risk group for purposes of these analyses.

All patients who received study drug and underwent CABG were included in the intent-to-treat analysis. The comparability of baseline characteristics in the two treatment groups was assessed for discrete variables by the Pearson g 2 test or Fisher's exact test. Sparse data usually led to the use of Fisher's exact test. For continuous variables, analysis of variance techniques with rank trans- formation were used. For outcome measures, the Coch- ran-Mantel-Haenszel test was used to compare the re- sponse for the two treatments. Control was imposed for stratum (center and risk group). When the data were too sparse to permit valid use of the Cochran-Mantel- Haenszel test, the strata were ignored and the incidences were compared by either the ~ test or Fisher s exact test. Additional logistic regression analyses were conducted to identify predictors of MI after adjusting for parameters of surgical risk. 43 All reported p values are two sided and compare the two treatment groups.

Results

A total of 862 patients were screened for enroll- ment in the study. Of these, 41 patients did not enter the study and were considered nonparticipants. The remaining 821 patients were stratified by risk group, allocated to treatment assignment, and received study medication (placebo, n = 418; acadesine, n = 403). The baseline characteristics of the two treatment groups were similar (Table I), and no significant differences in the extent of surgical exposure were detected (Table II). In addition, no statistically significant differences between treatment groups in the high-risk patients were noted for any of the baseline or surgical parameters. All patients were followed through discharge from the hospital; 84% of the discharged patients (placebo, 339/405; acadesine, 328/392) received a follow-up examination 4 to 5 weeks after the operation and 72% of patients completed the 6-month follow-up interval (placebo, 288/404; acadesine, 287/392). Data from only 59% of the patients (placebo, 237/402; acadesine, 231/391) were available at the 12-month follow-up survey.

Morbidity resulting from adverse cardiovascular outcomes. The incidence of MI was not statistically different between the two treatment groups according to any of the criteria for diagnosis (Table III). However, acadesine significantly reduced the incidence of MI in the prospectively identified high-risk patients (placebo, 19.7%; acadesine, 10.0%; p = 0.032) when MI was diagnosed by Q-wave or au-

Table I. Baseline characteristics*

Placebo Acadesine Characteristic (n = 418) (n = 403)

Sex ratio, M/F (%) 83/17 84/16 Mean age (yr)t 61.4 -4- 0.4 61.3 -4- 0.4 Mean weight (kg) 76.1 _+ 0.6 75.9 _+ 0.6 Cardiovascular history (%)

Stable angina 90 89 Previous MI 56 56 Hypercholesterolemia 53 55 Hypertension 43 41 Unstable angina:~ 31 30 Arrhythmias 12 12 Previous PTCA 8 7 Congestive hear t failure 3 5 Carotid bruit 5 4 Previous CABG 3 4 Valvular hear t disease 2 3

Risk group (%) Non-high risk 63 60 High risk 37 40

High-risk factor (%) Unstable angina§ 22 23 Age >70 yr 18 17 Previous CABG 3 4 Ejection fraction <30% 4 3 Acutely unsuccessful PTCA <1 0

Preoperative catheterization ->50% stenosis left main 18 16

coronary artery (%) ->70% stenosis of two or 93 94

more coronary arteries (%)

Mean No. of s tenosed ves- 5.9 _+ 0.1 5.8 _+ 0.1 selstt

Mean % ejection fraction 56 _+ 1 58 _+ 1 Mean L V E D P (mm Hg) 17 _ 1 17 - 1

MI, Myocardial infarction; PTCA, percutaneous transluminal coronary angioplasty; CABG, coronary artery bypass grafting; LVEDP, left ventricular end-diastolic pressure. *No differences were statistically significant. ?Mean -+ standard error. :~Any past history of unstable angina. § Unstable angina was defined as (1) new-onset angina or progressively worse angina (increased frequency, intensity, or duration) that was less responsive to therapy within 7 days of CABG surgery or (2) angina at rest unresponsive to maximal therapy. []Does not account for degree of stenosis.

topsy criteria (Table IV). Unstable angina repre- sented the most frequently occurring risk factor among the high-risk patients. Univariate analysis of the incidence of MI by risk factor showed that acadesine treatment significantly reduced the incidence of MI in those high-risk patients with unstable angina as compared with patients receiving placebo (p = 0.002). No statistically significant differences were noted between the two treatments for the other individual risk factors. No clear evidence

1 1 0 0 MenaschO et al. The Journal of Thoracic and


Table II. Surgicalprofiles* Placebo Acadesine

Procedure (n = 418) (n = 403)

Mean No. of vessels grafted 3.2 3.2 Type of graft (%)

IMA and vein 69 74 Vein only 19 17 IMA only 12 9

Mean duration of operation (min) Sternotomy to chest closure 197 202 Before bypass t 124 127 Bypass 97 96 Crossclamp:~ (n) 55 (384) 55 (371) Intermittent crossclamp (n) 32 (32) 29 (32) After bypass§ 45 47

Mean IV infusion (hr) 7.05 7.05 Cardioplegia composit ion (%)

Containing blood[[ 27 25 Containing no blood 66 67 No cardioplegia 7 8

IMA, Internal mammary artery; IV, intravenous. *No differences were statistically significant. tTime from the start of infusion of study drug to start of bypass, :~Not available for two patients and excludes those patients with intermittent crossclamping. §Time from end of bypass to chest closure. IIUse of warm blood cardioplegic solutions was discouraged.

could be found that treatment with acadesine showed any different effect on the basis of the number of vessels grafted or the type of cardioplegic solution used during bypass.

No significant between-group difference in the incidence of the combined adverse cardiovascular outcomes was observed (Table III). The rate for each of the individual adverse cardiovascular outcomes other than MI was small, with no statistically significant differences between the treatments. The difference in the reductions observed for the individual outcomes compared with the incidence of combined adverse cardiovascular outcomes may be explained by the fact that twice as many patients in the placebo group (16/81) had more than one adverse cardiovascular outcome as compared with the acadesine group (7/74).

The mean peak CK-MB level (78 + 11 ng/ml) observed for placebo-treated patients with a Q-wave MI was lower than expected as compared with reports of CK-MB levels measured with the same assay kit (Tandem-E CKMB II) in patients having CABG and subsequent MIs diagnosed by another modality (ECG, scintigraphy), z7̀ 40, 44, 45

Multivariate regression analyses were performed to evaluate the effect of acadesine on the incidence of MI (Q wave or autopsy) after adjusting for

Table III. Incidences of adverse cardiovascular outcomes in total study population

Placebo Acadesine (n = 418) (n = 403)

P Outcome variable No. * % No. % Value

M I t Q wave or CK-MB, 122/418

or autopsy Q wave and CK-MB, 27/402

or autopsy Q wave or autopsy 65/414

Combined adverse car- 81/418 diovascular outcomes:~

Cardiac death 7/418 CVA 7/418 Severe LV dysfunc- 14/418

?ion Life-threatening ar- 12/418

rhythmia

29.2 108/403 26.8 NS

6.7 20/385 5.2 NS

15.7 51/399 12.8 NS 19.4 74/403 18.4 NS

1.7 5/403 1.2 NS 1.7 5/403 1.2 NS 3.3 13/403 3.2 NS

2.9 9/403 2.2 NS

MI, Myocardial infarction; NS, not significant; CK-MB, creatine kinase myocardial band; CVA, cerebrovascnlar accident; LV, left ventricular. *Number of patients/number of patients evaluable. ?Diagnosis of a Q wave required a screening ECG and a sufficient number of postoperative ECGs to allow clinical diagnosis; diagnosis by CK-MB criteria required at least two postoperative CK-MB determinations. ~:MI (Q wave or autopsy), cardiac death, CVA, severe left ventricular dysfunction, or life-threatening arrhythmia.

Table IV. Incidence of MI in high-risk subgroup

Placebo Acadesine (n = 418) (n = 403)

P MI * No. 7" % No. % Value

High-risk subgroup 30/152 19.7 16/160 10.0 0.032 Unstable Angina 25/93 26.9 10/91 11.0 0.002 Age >70 yr 13/76 17.1 7/70 10.0 NS Previous CABG 3/12 25.0 1/18 5.6 NS EF < 3 0 % 3/16 18.8 1/12 8.3 NS Unsuccessful 1/1 100.0 0 0 -

PTCA

MI, Myocardial infarction; NS. not significant; CABG, coronary artery bypass grafting; EF, ejection fraction; PTCA, percutaneous transluminal coronary angioplasty. *Diagnosis by Q wave or autopsy. ?Number of patients/number of patients evaluable.

parameters of surgical risk, including type of cardioplegia used, gender, history of severe left ventricular dysfunction, previous MI, or stratification to the high-risk subgroup. When this model was fit to all patients (both high-risk and non-high risk), there was a significant interaction between acadesine and risk group (p = 0.028) (Table V). Acadesine significantly reduced the incidence of MI in high-risk patients (odds ratio = 0.44, p = 0.015) but not in non-high-risk patients (odds ratio = 1.12, p = 0.665).


MenaschO et al. 1 1 0 1

Table V. Logistic regression analysis for preoperative and operative risk factors*

All patients High-risk patients (n = 805) (n = 309)

Odds p Odds p Risk factor ratio Value ratio Value

Acadesine

Crystalline cardioplegia Blood cardioplegia Female Congestive heart

failure Cardioplegia used Previous MI Acadesine, high risk§ll High risklJ

0.44t 0.015 0.45 0.017 1.12:~ 0.665 0.81 0.635 0.72 0.624 0.71 0.362 0.76 0.647 0.93 0.793 0.78 0.548 1.14 0.801 0.95 0.939

1.20 0.756 1.14 0.873 1.33 0.183 1.86 0.089 0.39 0.028 1.60 0.092

MI, Myocardial infarction. *Patients had to be evaluable for MI (Q wave or autopsy) and all of the risk factors to be included in the analysis. tComparing acadesine to placebo in high-risk patients. :~Comparing acadesine to placebo in non-high-risk patients. §Interaction of acadesine and risk group. [fRisk factor not included in model for high-risk patients.

High-risk patients had a higher MI rate than non-high- risk patients (odds ratio = 1.60,p = 0.092); and in the former group, acadesine treatment was the only factor that significantly reduced the incidence of MI (odds ratio = 0.45, p = 0.017). Other risk factors tested did not have a significant effect. Table V also presents the results obtained when this model was used only in the high-risk subgroup, which were consistent with those obtained for all patients. An interesting finding from this analysis was that prospectively defined high-risk patients with a previous MI had a greater risk for the occurrence of MI associated with CABG (odds ratio = 1.86, p = 0.089) than did the overall population with a previous MI (odds ratio = 1.33, p = 0.183).

Mortality. There were 25 deaths through the 4- to 5-week foUow-up period. Hospital mortality was not statistically different between the two groups (placebo 3.4%, 13/418; acadesine 2.7%, 11/403). One death in the placebo group was reported for the period between hospital discharge and the 4- to 5-week follow-up visit. However, the number of deaths with acadesine treatment was significantly lower than with placebo (placebo 1.9%, 8/418; acadesine 0.2%, 1/403; p = 0.038) during the first 3 postoperative days. It is noteworthy that perioperative myocardial insufficiency (manifest as low output syndrome usually associated with MI, arrhythmia, or severe ventricular dysfunction, which were reported as adverse events) contributing to death was more

Table VI. Cause of death

Placebo Acadesine Cause of death (n = 418) (n = 403)

Cardiovascular 13 9 Myocardial insufficiency* 10 5 Stroke 0 2t Other:~ 3 2

Noncardiovascular 1 2 All causes 14 11

*Includes myocardial events such as MI, arrhythmia, or severe ventricular dysfunction, which were considered to be major contributory factors ultimately leading to death. tRespiratory complications were also present. :]:Placebo: two mesenteric infarctions and one pulmonary embolism; acadesine: two hemorrhages. Of the patients dying of hemorrhage, one had retrosternal bleeding; the other patient (a Jehovah's Witness who refused blood transfusion) died of edema of the lung and respiratory insufficiency resulting from blood loss.

prevalent in the placebo group than in the acadesine group (placebo 10/14; acadesine 5/11) (Table VI). The number of late deaths (5 to 52 weeks of follow-up) was similar for the two treatment groups (placebo 2/288; acadesine 2/287).

Adverse experiences and laboratory evaluations. Acadesine was very well tolerated. The overall adverse experience rates for the placebo and acadesine-treated groups were identical (82%). Atrial fibrillation was the most frequently reported adverse event in this study, but the rates were the same for both treatment groups (25%). No clinically significant differences were detected between acadesine and placebo treatment among other cardiovascular events: atrioventricular block, first- or second-degree (placebo 4.3%; acadesine 2.2%); heart block, third-degree (placebo 1.4%; acadesine 1.2%); and hypotension (placebo 15.3%; acadesine 20.3%). Hemodynamic measurements, including systolic and diastolic blood pressures and heart rate taken during the perioperative period, did not reveal any significant differences between the treatment groups.

With the exception of mild, transient elevation of serum uric acid levels at the end of the acadesine infusion (27% above baseline level) (Fig. 1), there were no differences in clinical laboratory measurements between treatment groups. Elevations in uric acid concentration were transient and without clinical sequelae.

Discussion

Acadesine did not significantly affect the incidence of MI, defined by the presence of ECG and enzyme criteria or by autopsy; however, acadesine

1102 Menasch~ et aL The Journal of Thoracic and

Cardiovascular Surgery

October 1995

[] Placebo [] Acadesine

30

% Change 20 from

Baseline (Mean +S.E.) 10

0

-10

-20 Post End End POD 1 Discharge

Protamine Infusion Infusion +8hr

Fig. 1. Percent change in uric acid concentration from baseline. POD, postoperative day; SE, standard error.

significantly reduced the incidence of MI, defined only by ECG and autopsy criteria, in the prospectively defined high-risk patient group (p = 0.032). Although the incidence of hospital mortality was not reduced, acadesine treatment resulted in a reduced proportion of cardiac related deaths, as compared with placebo, which produced a statistically significant reduction in the acute postoperative mortality (p = 0.038).

Acadesine and myocardial protection. Surgical and anesthetic practice is continually improving; however, the number of high-risk patients undergoing CABG is steadily increasing with an attendant rise in operative morbidity and mortality. Numerous attempts to preserve mYocardial viability by decreas- ing the effects of ischemia and reperfusion injury that occur during CABG have been made. Although many of them have been based on changes in the vehicle (crystalloid or blood), temperature (hypo- thermic or normothermic), or route or technique of delivery (antegrade or retrograde, intermittent or continuous) of cardioplegic solutions, or on appropriate alterations in reperfusate composition, others have relied on the supplementation of cardioplegic or reperfusion solutions with substrates or drugs. Among these additives has been adenosine, the cardioprotective effects of which have been known for many years. 46'47 Interest in adenosine is cur- rently being renewed, partially because of its purported role in mediating the protective effects of ischemic preconditioning. 48 However, the systemic

administration of adenosine is usually accompanied by unacceptable adverse events such as heart block, hypotension with reflex tachycardia, coronary steal, facial flushing, and dyspnea, which have limited its therapeutic use. The development of an adenosine- regulating agent such as acadesine may obviate the exogenous administration of adenosine, with its associated systemic adverse events, by enhancing local endogenous production or accumulation of adenosine. Although the mechanism by which acadesine may modulate local extracellular adenosine concentrations during periods of ischemia in animals has not been fully established, these cardioprotective effects are prevented when adenosine receptors are blocked or the formation of adenosine is suppressed, indicating that the ability of acadesine to enhance ischemia-induced adenosine levels is the basis of its beneficial effects. Modulation of local adenosine levels in ischemic myocardium by an adenosine-regulating agent should be without the systemic adverse effects associated with exogenous adenosine administration, because adenosine has an extremely short half-life in human blood (<0.6 second). Consequently, adenosine-regulating agents such as acadesine could have major clinical benefits in conditions in which it could prevent the progres- sion of myocardial ischemia to necrosis (i.e., MI), with its attendant morbidity and mortality.

Methodologic considerations. The incidence of perioperative MI varies greatly depending on the criterion used for the diagnosis. However, there is

The Journal of Thoracic and Cardiovascular Surger~ Volume 110, Number 4, Part 1

Menasch~ et al. 1 1 0 3

no consensus on the criterion by which a perioperative MI should be diagnosed in the setting of CABG. In the design of this study, we recognized that establishing the diagnosis of an MI in the perioperative period is more complicated than in the usual clinical setting in the ambulatory patient. The diagnosis of a perioperative MI by use of only the occurrence of new Q waves is associated with false- positive and false-negative results. False-positive results occur especially with new, inferior Q waves resulting from the unmasking of previously undetec- ted infarcts. 49 False-negative results occur with non- transmural MIs.

The design of the protocol allowed for analysis with two definitions used for the diagnosis of MI: (1) the presence of either a new Q wave on the ECG or elevations of CK-MB levels or a patho- logically confirmed myocardial necrosis at autopsy MI (Q wave or CK-MB, or autopsy); (2) the presence of both ECG changes and CK-MB elevations, or autopsy MI (Q wave and CK-MB, or

autopsy). With the more specific and clinically rigorous definition (Q wave and CK-MB, or autopsy), the incidence of MI was not significantly lower in the acadesine-treated group (5.2%) than in the placebo group (6.7%). Although these rates are compatible with those reported in the literature, diagnosis of a perioperative MI with the CK-MB data was considered to be unreliable in this study and therefore reduced the ability to diagnose MI by use of any definition that included a CK-MB criterion.

The unreliability of the CK-MB results is demonstrated in several ways. For patients with a Q wave who received placebo, the CK-MB concentrations were much lower than results reported in the literature for similar patients having CABG in whom the same assay (Tandem-E CKMB II) was used. 27' 40, 44, 45 CK-MB samples from the present study were subjected to freezing, prolonged storage, and international transport. The cited literature reports were all single center studies in which CK-MB values were determined on nonfrozen samples, which were pre- sumably assayed rapidly without being subjected to sample handling procedures with the potential for sample degradation. The CK-MB values determined during the validation of the CK-MB assay at the central laboratory for the present study on samples obtained from nonstudy patients, including some with Q-wave MIs after CABG, were dramat- ically higher than study patient samples and were compatible with those reported in the literature. 45

The logistics of handling the samples in this multinational study could account for the discrep- ancies between the lower than expected values reported here and the data obtained during the CK-MB assay validation and from the other literature reports in which the same assay kit was used. Additional evidence of unreliability of the CK-MB data comes from the fact that 50% of the patients with MIs confirmed anatomically at autopsy did not meet the CK-MB criteria for the diagnosis of MI; this observation is at variance with a study by Van Lente and associates. 5° Consequently, the presence of a new Q wave or evidence of MI on autopsy were adopted as the criteria for the diagnosis of MI in this study.

Efficacy of acadesine. Acadesine treatment did not significantly affect the incidence of MI in the overall population. The finding that acadesine significantly reduced the incidence of MI, with the Q-wave or autopsy definition used for the diagnosis, in high-risk patients (p = 0.032), however, is of potential clinical relevance, because Q-wave MIs have been shown to be associated with a worse long-term prognosis in the Coronary Artery Surgery Study. 15 We recognized before the start of the study that the high-risk patients would have a higher incidence of adverse outcomes and the establish- ment of a treatment effect in this group is not surprising. The lack of a statistically significant effect in the non-high-risk group may have resulted from a lower incidence of MIs in this subgroup, the occurrence of false-positive Q waves associated with CABG, which may preclude the detection of a true drug effect, 5° or the presence of a lesser or no drug effect. Multivariate analyses demonstrated that acadesine treatment had a significant effect in reducing the incidence of MI in the high-risk subgroup, whereas the interaction between acadesine and parameters of surgical risk did not have a significant effect on MI. The type of cardioplegic solution used had no significant effects on outcome, although the sample size was insufficient to make a definitive conclusion.

The incidence of combined adverse cardiovascular outcomes was highly dependent on the incidence of MIs because of the lower rates for the other individual adverse cardiovascular outcomes. Al- though the rates for the combined adverse cardiovascular outcomes were not different between treatments, there were potentially clinically meaningful reductions in the rates for cardiac death, cerebrovascular accident, life-threatening arrhythmia, and a

1 1 0 4 MenaschO et al. The Journal of Thoracic and

Cardiovascular Surgery

October 1995

minimal reduction for left ventricular dysfunction with acadesine as compared with placebo.

Results of the present study also demonstrate that although the overall hospital mortality rates were similar in the two groups, acadesine therapy was associated with a lower incidence of cardiac related events (i.e., low output syndrome, arrhythmia, severe left ventricular dysfunction) that accounted for the majority of deaths occurring over the first 3 postoperative days (placebo, 6/8; acadesine, 1/1). Consequently, mortality during this early postoperative period was lower in acadesine-treated, patients (placebo, 8/418; acadesine, 1/403; p = 0.038).

A similarly designed study has also been conducted in the United States. 51 The results of the U.S. study demonstrated that acadesine decreased the incidence of MI, but this decrease was not significant according to the prospectively defined criteria for the diagnosis of MI (Q wave or CK-MB, or autopsy). However, an additional analysis that used a more specific definition for MI (Q wave and CK-MB, or autopsy) showed a statistically significant reduction in the incidence of MI with acadesine treatment (p = 0.018).

Safety profile. This study also demonstrated that, for patients undergoing CABG, acadesine is safe when given at a dosage of 0.1 mg/kg per minute for 7 hours in association with administration in cardioplegic solutions at a final concentration of 5 /xg/ml. Systemic blood pressure and heart rate showed no difference in the acadesine-treated patients compared with the group receiving placebo. The overall adverse event profile for the acadesine- treated patients was essentially indistinguishable from that of the placebo recipients. Conduction defects of various types were reported less frequently with acadesine or at a similar rate to placebo. These results demonstrate that unlike adenosine, acadesine does not lower systemic blood pressure or produce conduction disturbances. The only potential safety concern was a mild, transient hyperuricemia, which was maximal at the end of the infusion and resolved shortly thereafter, with no clinical sequelae. The rise in uric acid was not unexpected because acadesine is metabolized via normal purine pathways of which the end product is uric acid.

Conclusions. This study has established that acadesine, given both intravenously during the perioperative period (at 0.1 mg/kg per minute over 7 hours) and as an adjunct to cardioplegic solutions

(at a final concentration of 5/xg/ml), is safe with an adverse event profile similar to that of placebo. Although the results of this study were not statistically significant for the reduction of MI or hospital mortality in the overall population, the reduced incidence of MI in the high-risk subgroup, according to secondary Q-wave or autopsy criteria, and of cardiac related deaths in the acute postoperative period (first 3 postoperative days) suggests that acadesine may provide some added myocardial protection during CABG.

R E F E R E N C E S 1. Naunheim KS, Fiore AC, Wadley J J, et al. The

changing profile of the patient undergoing coronary artery bypass grafting. J Am Coll Cardiol 1988;11: 494-8.

2. Christakis GT, Ivanov J, Weisel RD, Birnbaum PL, David TE, Salerno TA. The changing pattern of coronary artery bypass surgery. Circulation 1989; 80(Suppl):1151-61.

3. Jones E, Weintraub W, Craver J, Guyton RA, Cohen CC. Coronary bypass surgery: Is the operation different today? J THORAC CARDIOVASC SURG 1991; 101:108- 15.

4. Mullany C, Darling GE, Pluth JR, et al. Early and late results after isolated coronary artery bypass surgery in 159 patients aged 80 years and older. Circulation 1990;82(Suppl):IV229-36.

5. Weintraub WS, Craver JM, Cohen CL, Jones EL, Guyton RA. Influence of age on results of coronary artery surgery. Circulation 1991;84(Suppl):III226-35.

6. Grover FL, Hammermeister KE, Burchfiel C. Initial report of the Veterans Administration preoperative risk assessment study for cardiac surgery. Ann Thorac Surg 1990;50:12-28.

7. Christakis GT, Weisel RD, Fremes SE, et al. Coro- nary artery bypass grafting in patients with poor ventricular function. J THORAC CARDIOVASC SURG 1992;103:1083-92.

8. Higgins TL, Estafanous FG, Loop FD, Beck GJ, Blum JM, Paranandi L. Stratification of morbidity and mortality outcome by preoperative risk factors in coronary bypass patients. JAMA 1992;267:2344-8.

9. Davis P, Parascandola SA, Miller CA, et al. Mortality of coronary artery bypass grafting before and after the advent of angioplasty. Ann Thorac Surg 1989;47: 493-8.

10. Hartz AJ, Kuhn EM, Pryor DB, et al. Mortality after coronary angioplasty and coronary artery bypass surgery (The National Medicare Experience). Am J Cardiol 1992;70:179-85.

11. ACC/AHA Task Force Report. ACC/AHA Guide- lines and indications for coronary artery bypass graft


Menasch~ et al. 1 1 0 5

surgery: a report of the American College of Cardi- ology/American Heart Association Task Force on assessment of diagnostic and therapeutic cardiovascular procedures (subcommittee on coronary artery bypass graft surgery). J Am Coll Cardiol 1991;17:543-89.

12. Schaff H, Gersh B, Fisher L, et al. Detrimental effect of perioperative myocardial infarction on late survival after coronary artery bypass: report from the Coro- nary Artery Surgery Study--CASS. J THORAC CAR- DIOVASC SURG 1984;88:972-81.

13. Val P, Pelletier L, Hernandez M, et al. Diagnostic criteria and prognosis of perioperative myocardial infarction following coronary bypass. J THORAC CAR- DIOVASC SURG 1983;86:878-86.

14. Fennel W, Chua KG, Cohen L, et al. Detection, prediction, and significance of perioperative myocardial infarction following aorta-coronary bypass. J THORAC CARDIOVASC SURG 1979;78:244-53.

15. Chaitman BR, Alderman EL, Sheffield LT, et al. Use of survival analysis to determine the clinical significance of new Q waves after coronary bypass surgery. Circulation 1983;67:302-9.

16. Force T, Hibberd P, Weeks G, et al. Perioperative myocardial infarction after coronary artery bypass surgery. Circulation 1990;82:903-12.

17. Seitelberger R, Zwolfer W, Huber S, et al. Nifedipine reduces the incidence of myocardial infarction and transient ischemia in patients undergoing coronary bypass grafting. Circulation 1991;83:460-8.

18. Slogoff S, Keats AS. Further observations of perioperative myocardial ischemia. Anesthesiology 1986;65: 539-42.

19. Gardner M J, Johnstone DE, Lalonde L, et al. Peri- operative myocardial infarction with coronary artery surgery: diagnosis, incidence and consequences. Can J Cardiol 1987;3:336-41.

20. Devine JE, Wiens RD, Halstead JM, Codd JE. Quan- titation of CK-MB release: diagnostic utility in coronary artery bypass grafting. Clin Chim Acta 1986;156: 145-50.

21. Loeb HS, Reid R, Pifarr6 R, Gunnar RM, Scanlon PJ. Usefulness of postoperative enzymes for diagnosis of perioperative myocardial infarction following aortocoronary bypass. Acute Care 1985;11:40-7.

22. Balderman SC, Bhayana JN, Steinbach J J, Masud AR, Michalek S. Perioperative myocardial infarction: a diagnostic dilemma. Ann Thorac Surg 1980;30: 370-7.

23. Knight AA, Hollenberg M, London M J, Mangano DT. Perioperative myocardial ischemia: importance of the perioperative ischemic pattern. Anesthesiology 1988;68:681-8.

24. Roberts AJ, Spies SM, Lichtenthal PR, Moran JM, Sanders JH, Michaelis LL. Changes in left ventricular performance related to perioperative myocardial in-

farction in coronary artery bypass graft surgery. Ann Thorac Surg 1983;35:516-24.

25. Farah SY, Moss DW, Ribiero P, Oakley CM, Saps- ford RN. Interpretation of changes in the activity of creatine kinase MB isoenzyme in serum after coronary artery bypass grafting. Clin Chim Acta 1984;141: 219-25.

26. Gray RJ, Matloff JM, Conklin CM, et al. Periopera- tive myocardial infarction: late clinical course after coronary artery bypass surgery. Circulation 1982;6: 1185-9.

27. Chapelle JP, E1 Allaf M, Larbuisson R, Limet R, Lamy M, Heusghem C. The value of serum CK-MB and myoglobin measurements for assessing perioperative myocardial infarction after cardiac surgery. Scand J Clin Lab Invest 1986;46:519-26.

28. Burns RJ, Gladstone PJ, Tremblay PC, et al. Myocar- dial infarction determined by technetium-99m pyrophosphate single-photon tomography complicating elective coronary artery bypass grafting for angina pectoris. Am J Cardiol 1989;63:1429-34.

29. Cheng DC, Chung F, Burns R J, Houston PC, Feindel CM. Postoperative myocardial infarction documented by technetium pyrophosphate scan using single-photon emission computed tomography: significance of intraoperative myocardial ischemia and hemodynamic control. Anesthesiology 1989;71: 818-26.

30. Griesmacher A, Grimm M, Schreiner W, Muller MM. Diagnosis of perioperative myocardial infarction by considering relationship of postoperative electrocardiogram changes and enzyme increases after coronary bypass operation. Clin Chem 1990; 36:883-7.

31. Kato NS, Buckberg GD. New intraoperative strategies for myocardial protection. Curr Opinion Cardiol 1992;7:959-67.

32. Slogoff S, Keats A. Does perioperative myocardial ischemia lead to postoperative myocardial infarction? Anesthesiology 1985;62:107-14.

33. Downey JM, Liu GS, Thornton JD. Adenosine and the anti-infarct effects of preconditioning. Cardiovasc Res 1993;27:3-8.

34. Gruber HE, Hoffer ME, McAllister DR, et al. In- creased adenosine concentration in blood from ischemic myocardium by AICA riboside: effects on flow, granulocytes and injury. Circulation 1989;80:1400-11.

35. Mullane K. Acadesine: the prototype adenosine regulating agent for reducing myocardial ischemic injury. Cardiovasc Res 1993;27:43-7.

36. Mullane K, Young M. Acadesine: prototype adenosine regulating agent for treating myocardial ischemia-reperfusion injury. Drug Dev Res 1993;28:336-43.

37. Bolling SF, Groh MA, Mattson AM, Grinage RA, Gallagher KP. Acadesine (AICA-riboside) improves

1 1 0 6 Menasch~ et aL The Journal of Thoracic and


postischemic cardiac recovery. Ann Thorac Surg 1992; 54:93-8.

38. Vinten-Johansen J, Nakanishi K, McGee S, Zhao ZQ. Acadesine improves surgical myocardial protection with blood cardioplegia in ischemically injured hearts [Abstract]. Circulation 1992;86:(Suppl):I104.

39. Leung JM, Stanley T, Mathew J, et al. An initial multicenter, randomized controlled trial on the safety and efficacy of acadesine in patients undergoing CABG surgery. Anesth Analg 1994;78:420-34.

40. Van Lente F, McHugh AM, Galen RS. Changes in serum CK-MB mass after coronary artery bypass surgery. Clin Biochem 1987;20:333-7.

41. Chapelle JP, E1 Allaf M, Heusghem C. Serum CK-MB after myocardial infarction and coronary bypass surgery: superiority of mass concentration measurement over determination of catalytic activity. Adv Clin Enzymol 1987;5:94-102.

42. Lachin JM. Sample size determinations for r x c comparative trials. Biometrics 1977;33:315-24.

43. Hosmer DW, Lemeshow S. Applied logistic regression. New York: John Wiley, 1989.

44. Gulbis B, Unger P, Lenaers A, Desmet JM, Ooms HA. Mass concentration of creatine kinase MB isoenzyme and lactate dehydrogenase isoenzyme 1 in diagnosis of perioperative myocardial infarction after coronary bypass surgery. Clin Chem 1990;36:1784-8.

45. Hossein-Nia M, Johnston A, Holt DW. Performance of the Tandem-E CKMB assay for the measurement of creatine kinase MB isoenzyme concentration. Commun Lab Med 1993;2:7-11.

46. Engler R. Consequences of activation and adenosine mediated inhibition of granulocytes during myocardial ischemia. Fed Proc 1987;46:2407-12.

47. Berne RM. Adenosine--a cardioprotective and therapeutic agent. Cardiovasc Res 1993;27:2.

48. Mullane, K. Myocardial preconditioning: part of the adenosine revival. Circulation 1992;85:845-7.

49. Bassan MM, Oatfield R, Hoffman I, Matloff J, Swan HJ. New Q waves after aortocoronary bypass surgery. N Engl J Med 1974;290:349-53.

50. Van Lente F, Martin A, Ratliff NB, Kazmierczak SC, Loop FD. The predictive value of serum enzymes for perioperative myocardial infarction after cardiac operations. J THORAC CARDIOVASC SURG 1989;98:704-10.

51. Mangano DT, for the Multicentre Study of Periop- erative Ischemia (McSPI) Research Group. Multicen- tre study of acadesine for the prevention of myocardial infarction and other adverse cardiovascular outcomes associated with coronary artery bypass graft surgery. Eur Heart J 1994;15(abs suppl):66.

Appendix The following institutions (listed in alphabetical order)

and principal investigators participated in the Interna- tional Acadesine Study:

Medical center Aberdeen Royal Infirmary, U.K.

Academisch Ziekenhuis, Maastricht, The Netherlands

Akademiska Sjukhuset, Uppsala, Sweden

C.H.U. Henri M6ndor, Cr6teil, France C.H.U. Rangueil, Toulouse, France Freeman Hospital, Newcastle, U.K. Gasthuisberg University Clinic, Leuven,

Belgium Glasgow Royal Infirmary, Glasgow, U.K. H6pital Augustin Morvan, Brest, France H6pital Broussais, Paris, France H6pital Lariboisi6re, Paris, France Institut Arnault Tzanck, Nice, France Klinikum Gro/3hadern, Munich, Germany Laval Hospital, Ste. Foy, Canada Middlesex Hospital, London, U.K. Montreal Heart Institute, Montreal,

Canada National Hospital, Oslo, Norway Notre Dame Hospital, Montreal, Canada Royal Infirmary of Edinburgh, U.K.

St. Thomas' Hospital, London, U.K. Toronto General Hospital, Toronto,

Canada Vancouver General Hospital, Vancouver,

Canada University Hospital, Lund, Sweden

University of Alberta Hospitals, Edmonton, Canada

University of Freiburg, Freiburg, Germany

University of Heidelberg, Heidelberg, Germany

University of Ottawa Heart Institute, Ottawa, Canada

Principal investigator

R. Jeffrey, J. Cockburn

O. Penn

H. E. Hansson, S. Thelin

D. Loisance G. Fournial A. Hedley Brown W. Flameng

W. Dimitri J.-A, Barra J.-P. Couetil P. Menasch6 P. Lena T. Fischlein J. Metras W. Pugsley L. Pelletier

O. Geiran N. Poirier C. Campanella,

G. Sutherland G. Venn R. Weisel,

T. David W. R. E. Jamieson,

D. Ansley J. Kugelberg,

S. Steen A. Koshal,

E. Gelfand R. Seitelberger,

R. Fasol S. Hagl

P. Bedard, W. Keon

The central analysis centers participating in the study were as follows: ECG Analysis: Queen Elizabeth Hospital, University of Birmingham, U.K.; M. K. Davies, S. J. S. Virk, H. Hamdan, and K. Kullar. Dutch Heart Scan Services, Maasdam, The Netherlands; H. J. Ritsema van Eck. CK-MB Laboratory: St. George's Hospital Medical School, London, U.K.; D. Holt.

acadesine: a new drug that may improve myocardial … · 2016. 12. 12. · however, acadesine re-...

Documents