OUP user menu

High-dose intracoronary adenosine for myocardial salvage in patients with acute ST-segment elevation myocardial infarction

Walter Desmet , Jan Bogaert , Christophe Dubois , Peter Sinnaeve , Tom Adriaenssens , Christos Pappas , Javier Ganame , Steven Dymarkowski , Stefan Janssens , Ann Belmans , Frans Van de Werf
DOI: http://dx.doi.org/10.1093/eurheartj/ehq492 867-877 First published online: 31 December 2010

Abstract

Aims Previous studies have suggested that intravenous administration of adenosine improves myocardial reperfusion and reduces infarct size in ST-elevation myocardial infarction (STEMI) patients. Intracoronary administration of adenosine has shown conflicting results.

Methods and results In a prospective, single-centre, double-blind, placebo-controlled clinical study, we assessed whether selective intracoronary administration of adenosine distal to the occlusion site immediately before initial balloon inflation results in myocardial salvage and decreased microvascular obstruction (MVO) as assessed with cardiac magnetic resonance imaging (MRI). Using a combination of T2-weighted and contrast-enhanced sequences, myocardial salvage index (MSI) was defined as the percentage of the area at risk that did not become necrotic. We randomized 112 patients presenting with STEMI within 12 h from symptom onset to selective intracoronary administration of adenosine 4 mg or matching placebo. In 100/110 (91%) patients receiving study drug, MRI was performed on Days 2–3. No significant difference in MSI was found between adenosine- and placebo-treated patients: 41.3% (20.8, 66.7) vs. 47.8% (39.8, 60.9) [median (Q1, Q3)] (P = 0.52). The extent of MVO was comparable in both groups, with a trend favouring the placebo group: 2.4 g (0.0, 6.8) vs. 5.9 g (0.0, 12.8) after adenosine (P = 0.07). TIMI flow grade, TIMI frame count, myocardial blush grade, and ST-segment resolution after primary percutaneous coronary intervention were similar between groups. After 4 months, infarct size was similar in both treatment groups.

Conclusion We found no evidence that selective high-dose intracoronary administration of adenosine distal to the occlusion site of the culprit lesion in STEMI patients results in incremental myocardial salvage or a decrease in microvascular obstruction.

Clinical Trial Registration Information: ClinicalTrials.gov number, NCT00284323.

  • Myocardial infarction
  • Adenosine
  • Primary PCI
  • Magnetic resonance imaging

Introduction

Over the past decades, primary percutaneous coronary intervention (primary PCI) has emerged as the best treatment modality for evolving acute myocardial infarction with ST-segment elevation (STEMI).1 However, despite successful restoration of epicardial flow, impaired myocardial reperfusion is frequently observed after primary PCI, resulting in larger infarct size and increased mortality.26 Embolization of atherothrombotic material,7 vasoconstriction, platelet aggregation, neutrophil adhesion, myocardial oedema, and cardiomyocyte necrosis all contribute to microvascular dysfunction upon reperfusion. The endogenous nucleoside adenosine may improve myocardial perfusion and protect against reperfusion injury by a potent vasodilatory effect and possibly anti-inflammatory and platelet inhibition properties.8,9

Large randomized trials have suggested that intravenous administration of adenosine during intervention for STEMI may reduce infarct size, especially at higher doses.10,11 As dosing of intravenous adenosine is limited by systemic side effects, including hypotension, bradycardia, and chest pain, it is conceivable that intracoronary administration may result in higher local concentrations with less systemic side effects. An intracoronary bolus of adenosine 30–60 µg is recommended for prevention and treatment of no-reflow by the ESC guidelines for the treatment of STEMI (class IIb; level of evidence C).12 A small open-label clinical study has shown that intracoronary administration of 4 mg of adenosine is safe in patients with STEMI, with favourable effects on TIMI flow grade, incidence of no-reflow, left ventricular function, and clinical course.13 Improved myocardial reperfusion as assessed by ST-segment elevation resolution after primary PCI was observed in a cohort of 79 patients treated with intracoronary adenosine compared with 200 historical controls14 and in a randomized trial of 51 patients.15

The aim of this study was to investigate the effect of high-dose (4 mg) selective intracoronary administration of adenosine immediately before initial balloon inflation on microvascular and myocardial salvage in an unselected population of patients with STEMI. In the design of the present study, we attempted to address some of the methodological limitations of earlier trials. First, we used magnetic resonance imaging (MRI) to assess the primary endpoint, myocardial salvage, because it is less influenced by differences in the extent of ischaemic myocardium at baseline between the two treatment groups (Figure 1).1618 Second, all study personnel remained blinded to treatment assignment. Third, we administered the study drug sub-selectively to the ischaemic myocardium. Fourth, we used a very high dose of adenosine, and last, we administered the intracoronary adenosine before initial balloon inflation.

Figure 1

Example of the two extremes of myocardial salvage. Cardiac short axis T2-weighted (A and C) and contrast-enhanced magnetic resonance imaging (MRI) (B and D). Upper row: myocardial oedema in the anteroseptal LV wall (arrowheads) (A), without evidence of abnormal myocardial enhancement on contrast-enhanced images (B), representing an aborted myocardial infarction. Lower row: extensive myocardial oedema in the anterior and septal left ventricular (LV) wall (arrowheads) (C). On post-contrast enhanced MRI (D), transmural myocardial enhancement (arrowheads) of the anterior and septal LV wall is seen, involving nearly the entire area of myocardial oedema (arrowheads). Note the presence of an important area of microvascular obstruction (MVO)(*).

Methods

Population

The ‘beneficial effect of intracoronary adenosine on microvascular and myocardial SALVAGE in patients with acute myocardial infarction’ study was a single-centre, prospective, randomized, double-blind, placebo-controlled clinical trial with blinded evaluation of endpoints. All consecutive patients presenting to the University Hospitals of Leuven suspected of an acute STEMI and candidates for primary PCI were eligible for participation. Inclusion criteria were symptoms of chest pain suggestive of myocardial ischaemia for at least 20 min, a time from onset of symptoms of <12 h, and an ECG showing ST-segment elevation of >0.1 mV in two or more limb leads or >0.2 mV in two or more contiguous precordial leads, or presumed new left bundle-branch block. Exclusion criteria were contra-indication to heparin, low-molecular-weight heparin or clopidogrel, anticipated difficult vascular access, cardiogenic shock, inability to give informed consent, high-grade atrioventricular block, severe asthma, treatment with theophylline, glibenclamide, or dipyridamole, prior coronary artery surgery, and participation in any investigational drug or device study within the past 30 days. After coronary angiography, patients were randomized when PCI was indicated. The study was approved by the medical ethics committee of our institution.

Randomization and treatment

After coronary angiography, eligible patients were randomized (1:1) to a high-dose bolus injection of intracoronary adenosine (4 mg in 5 mL of 0.9% NaCl) or placebo (5 mL of 0.9% NaCl). Study drug was constituted by the hospital pharmacy in numbered vials according to a computer-generated randomization list in blocks of four that was stratified according to the time from symptom onset to primary PCI (<4 vs. >4 h). This list was kept in a sealed envelope in the hospital pharmacy. Vials containing adenosine and those containing placebo had an identical appearance. Hence, all study personnel was blinded for treatment allocation until the study had been completed and all analyses had been performed.

After crossing the obstruction of the infarct-related coronary artery with a long guide wire, an over-the-wire balloon (Maverick 1.5 mm × 12 mm, Boston Scientific, Natick, MA, USA) was positioned at the level of the obstruction. The guide wire was removed and a small quantity of diluted contrast medium was carefully injected through the central lumen of the balloon catheter to confirm positioning of the catheter tip downstream of the obstruction and to assess patency of the distal vessel. Consequently, the study drug solution was injected by hand through the central lumen of the balloon catheter into the distal vascular bed over 1min. The guide wire was then reinserted through the balloon catheter and advanced to a distal position, and the balloon was inflated. After deflation of the balloon, the procedure was continued per operator preference. As the study was conceived and started in 2006, neither thrombus aspiration, nor the use of any other additional device except coronary balloons and stents was allowed per protocol.

All patients received aspirin (≥250 mg), heparin (5000 IU), and clopidogrel (600 mg) after confirmation of ST-segment elevation on the first ECG. Before commencing primary PCI, patients received a bolus of the glycoprotein IIb/IIIa inhibitor abciximab (0.25 mg/kg), followed by a 12 h infusion. Additional heparin was administered to maintain an activated clotting time of 200–250 s. The standard treatment after primary PCI included aspirin, clopidogrel, β-blockers, lipid-lowering agents, and angiotensin-converting enzyme inhibitors.

Endpoints and definitions

The primary endpoint was myocardial salvage, defined as the percentage of the area at risk (AAR), which was not necrotic on MRI on Days 2–3. Area at risk was assessed by T2-weighted images, while necrosis was assessed by late enhancement. Microvascular obstruction (MVO) on MRI at Days 2–3, expressed as a percentage of the AAR, was a major secondary endpoint.

Other secondary endpoints were TIMI flow grade, TIMI frame count and myocardial blush grade at the end of the primary PCI, ST-segment resolution on the ECG after PCI, major adverse cardiac events in-hospital, at 30 days and at 1 year, recovery of LV-function as assessed using MRI at 4 months, and evolution of cardiac markers in the first 24 h.

In a pre-specified analysis, all endpoints were assessed in patients presenting within 4h of symptom onset and in those presenting later.

Cardiac magnetic resonance imaging

Cardiac MRI studies were performed on Days 2–3 and after 4 months. Detailed methodology of image acquisition and analysis has been reported earlier.18

Angiographic analysis

Coronary angiograms obtained before and after primary PCI were reviewed by two experienced observers blinded to treatment allocation and clinical data. On the initial angiogram and on the angiogram after stenting, TIMI flow grade was assessed.19 In addition, TIMI frame count and myocardial blush grade were assessed after stenting, as described previously.20,21 Distal embolization was defined as a filling defect distal to the culprit lesion on the angiogram at any time point after the first balloon inflation.

Electrocardiographic analysis

ECGs were analysed quantitatively in the ECG Core Laboratory of the Katholieke Universiteit Leuven, using a custom-made in-house computer assisted program.22

Biomarkers reflecting infarct size

Serum creatine kinase (CK), myocardial band of CK (CK-MB), and troponin I measurements were collected in all patients during hospitalization. Blood was collected on admission, just before and after primary PCI, and at 8, 16, and 24 h after revascularization.

Follow-up

Follow-up data at 30 days and 1 year after primary PCI were collected from hospital records and telephone interviews.

Acute side effects

Acute side effects were registered by the interventional cardiologist during and directly after infusion of the bolus intracoronary adenosine or placebo.

Sample size and statistical analysis

The study was powered to detect a difference of 20% in the primary endpoint between the two groups (common STD 30%) with 90% power at a significance level of 5% using a two-tailed t-test, requiring 100 MRIs and assuming 10% of patients for whom no adequate MRI can be done. Data were analysed according to the intent-to-treat principle. Missing data for the primary and major secondary endpoint were imputed according to a worst-case scenario. These analyses yielded similar results to the non-imputed analyses, hence only the latter are presented here. Categorical variables are presented as frequency values and proportions and were compared by the Cochran–Mantel–Haenszel test stratified by time to primary PCI. Continuous normally distributed variables are presented as mean values and standard deviations. For skewed distributed variables, median values with interquartile range (IQR) are shown. Comparisons between the groups were done by means of an analysis of variance (ANOVA) using randomized treatment and stratum as factors in the model. For non-normally distributed variables, a Van Elteren test was also performed, but was in all cases found to be consistent with the results of the ANOVA. Therefore, only results obtained using the ANOVA are presented. For the primary and major secondary endpoints, the average treatment effect was estimated using the ANOVA and presented along with its 95% confidence interval. Subgroup analyses were performed for myocardial salvage index (MSI) and MVO by age, gender, time to PCI, diabetes, baseline TIMI, final TIMI, and infarct location. All tests were two-sided and assessed at the 5% significance level. In view of the exploratory nature of this study, no correction was applied for multiple testing. Statistical analysis was performed using the SAS/STAT® software, version 9.2 of the SAS system for Windows. The authors had full access to the data and take responsibility for its integrity. All authors have read the manuscript and agree to its content.

Results

Between 17 February 2006 and 20 February 2009, 112 patients diagnosed with STEMI met the inclusion criteria and gave informed consent for participation in the study. After initial angiography, 110 patients underwent primary PCI and were randomized to intracoronary high-dose adenosine (n = 56) or matching placebo (n = 54) (Figure 2). Patients randomized to adenosine had a higher heart rate on admission when compared with patients randomized to placebo (76.6 ± 18.2 vs. 69.3 ± 18.3 b.p.m., P = 0.04; Table 1), and patients randomized to intracoronary administration of placebo more often had a medical history of diabetes mellitus (16.7 vs. 3.6%, P = 0.02) and previous PCI (11.1 vs. 1.8%, P = 0.05), as well as a longer stent length. TIMI flow 0 or 1 was present in 45 of the 56 patients (80.4%) randomized to intracoronary adenosine and in 34 of the 54 patients (63.0%) randomized to placebo (P = 0.044). All other baseline and procedural characteristics did not significantly differ between the two treatment groups.

View this table:
Table 1

Baseline characteristics

Adenosine (n = 56)Placebo (n = 54)
Age (years)61.4 (12.3)60.6 (10.8)
BMI (kg/m2)26.7 (3.9)27.1 (3.6)
Male gender44 (78.6)46 (85.2)
Heart rate (b.p.m.)76.6 (18.2)69.3 (18.3)
Systolic blood pressure (mmHg)131.8 (23.8)135.5 (25.8)
Diastolic blood pressure (mmHg)76.4 (15.2)78.5 (15.2)
Current smoker25 (44.6%)29 (53.7%)
History
 Myocardial infarction0 (0.0)1 (1.9)
 Diabetes mellitus2 (3.6)9 (16.7)
 PCI1 (1.8)6 (11.1)
 Family history20 (35.7)24 (44.4)
 Hypertension22 (39.3)22 (40.7)
 Angina pectoris9 (16.1)12 (22.2)
 Hypercholesterolaemia40 (71.4)33 (61.1)
 Ischaemic time (median, IQR), min215 (150, 296)193 (150, 305)
Angiographic
 Infarct-related vessel
  RCA22 (39.3%)29 (53.7%)
  LAD27 (48.2%)18 (33.3%)
  Cx7 (12.5%)7 (13.0%)
TIMI flow before PCI
 036 (64.3)28 (51.9)
 19 (16.1)6 (11.1)
 26 (10.7)13 (24.1)
 35 (8.9)7 (13.0)
 0/145 (80.4)34 (63.0)
 2/311 (19.6)20 (37.0)
 Stent diameter, mm3.2 (0.49)3.2 (0.42)
 Stent length, mm18.3 (5.52)21.6 (5.88)
 GPIIb/IIIa-inhibitor during PCI51 (91.1%)50 (94.3%)
  • PCI, percutaneous coronary intervention; RCA, right coronary artery; LAD, left anterior descending; Cx, circumflex artery; TIMI, thrombolysis in myocardial infarction.

Figure 2

Flow diagram: randomization, treatment, and MRI. OTW indicates over-the-wire; i.c., intracoronary; MRI, magnetic resonance imaging.

Magnetic resonance imaging results

Of the 110 randomized patients, 10 did not undergo MRI during the initial hospitalization: 1 patient was in hypovolemic shock, 1 in cardiogenic shock, 1 patient deceased 12 h after admission, and 7 refused MRI because of claustrophobia. As a result, cardiac MRI was performed on the second or third day after primary PCI in 51 patients who had received intracoronary adenosine and 49 patients who had received placebo during PCI (Table 2). Oedema volume as a measure for the AAR was 43.0 g (22.8, 54.4) in the patients randomized to adenosine and 40.3 g (19.7, 55.8) in the patients randomized to placebo (P = 0.87). Infarct volume was 22.0 g (6.7, 38.8) in the adenosine-treated patients vs. 19.9 g (10.4, 29.0) in the placebo-treated patients (P = 0.26), or—expressed as a percentage of the left ventricular mass—18.1% (7.6, 29.3) vs. 16.1% (10.1, 21.3) (P = 0.13). This resulted in a MSI of 41.3% (20.8, 66.7) in the patients randomized to intracoronary adenosine and 47.8% (39.8, 60.9) in the patients randomized to intracoronary placebo (P = 0.52) (Figure 3). Microvascular obstruction was observed in 37 of the 51 adenosine-treated patients (72.5%) when compared with 33 of the 49 (67.5%) placebo-treated patients (P = 0.60). The extent of early MVO was 5.9 g (0.0, 12.8) in the patients randomized to adenosine and 2.4 g (0.0, 6.8) in the patients randomized to placebo (P = 0.07). This lack of treatment effect was present in both strata, hence irrespective of time from symptom onset. Other MRI-analyses were comparable between the two treatment groups, with the exception of a lower ejection fraction in the adenosine-treated patients.

View this table:
Table 2

Magnetic resonance imaging results at Days 2–3

Adenosine (n = 51)Placebo (n = 49)P-value
Myocardial evaluation (median, IQR)
 Area at risk (g)43.0 (22.8, 54.4)40.3 (19.7, 55.8)0.87
 Microvascular obstruction present37/51 (72.5%)33/49 (67.3%)0.60
 Microvascular obstruction (g)5.9 (0.0, 12.8)2.4 (0.0, 6.8)0.07
 Infarct volume (g)22.0 (6.7, 38.8)19.9 (10.4, 29.0)0.26
 Infarct volume (% LV mass)18.1 (7.6, 29.3)16.1 (10.1, 21.3)0.13
 Myocardial salvage index (%)41.3 (20.8, 66.7)47.8 (39.8, 60.9)0.52
LV function (mean, SD)
 LV EDV (mL)159.0 (36.5)159.1 (40.0)0.97
 LV ESV (mL)83.4 (26.0)77.6 (30.2)0.32
 SV (mL)75.6 (19.6)80.3 (18.7)0.23
 EF (%)47.6 (9.5)51.3 (8.7)0.048
 LV mass (g)121.2 (34.8)124.0 (29.5)0.65
  • MRI, cardiac magnetic resonance imaging; LV EDV, left ventricular end-diastolic volume; LV ESV, left ventricular end-systolic volume; SV, stroke volume; EF, ejection fraction.

Figure 3

Box plot showing myocardial salvage index (primary endpoint). Box plot shows median and interquartile range. Whiskers are drawn at (Q3+1.5*IQR, Q1−1.5*IQR). Q1, Q3 = first and third quartile, IQR = Q3 − Q1. ‘+' sign indicates mean value. Small box indicates outlier.

At 4 months, ejection fraction slightly increased in both treatment groups despite a small increase in end-systolic LV volumes, primarily due to an increase in end-diastolic LV volumes (Table 3). Infarct volumes had roughly halved in both treatment groups, irrespective of symptom duration.

View this table:
Table 3

Magnetic resonance imaging results at 4 months

Adenosine (n = 46)Placebo (n = 44)P-value
LV function (mean, SD)
 LV EDV (mL)171.0 (48.8)168.3(44.0)0.81
 LV ESV (mL)87.1 (32.1)78.9 (33.6)0.26
 SV (mL)87.0 (24.0)89.4 (19.3)0.62
 EF (%)50.5 (9.9)54.4 (9.4)0.07
 LV mass (g)107.6 (30.5)110.1 (26.9)0.63
 Infarct volume (g)11.5 (4.0, 25.0)8.9 (4.5, 13.9)0.20
 Infarct volume (% LV mass)9.9 (4.8, 20.0)7.6 (4.5, 11.8)0.15

Additional analyses, adjusted for baseline TIMI, diabetes, time to PCI, and infarct location, were performed for the main endpoints of interest and yielded results similar to the unadjusted analyses. Subgroup analyses for MSI revealed a statistically significant interaction between treatment and baseline TIMI flow grade (0/1 vs. 2/3) (P = 0.0056), whereby a statistically significant treatment benefit of adenosine was found for patients with TIMI 2/3 at baseline (P = 0.0486). A worsening effect of treatment was found in the group of patients with TIMI flow grade 0/1 at baseline, but this difference was not statistically significant (P = 0.0676). These results could not be reproduced for MVO (Figure 4).

Figure 4

Forest plots of subgroup analyses for myocardial salvage index and microvascular obstruction.

ST-segment resolution

Two patients had a presumed new left bundle branch block on the qualifying electrocardiogram, and of the electrocardiograms of three patients, one was of insufficient quality to allow for reliable analysis. Hence, in 105 of the 110 patients (95.5%), residual ST-segment deviation could be assessed on the electrocardiogram after primary PCI. The median time between the completion of the primary PCI procedure to this electrocardiogram was 31 min (IQR, 19–41 min). ST-segment elevation resolution >70% was present in 24 of the 53 patients (45.3%) randomized to intracoronary adenosine and in 21 of the 52 patients randomized to intracoronary placebo (40.4%) (P = 0.45; Figure 5). Median ST-segment resolution was 65.5% (IQR, 46.9–77.4%) in the adenosine-treated patients when compared with 60.1% (IQR, 32.3–82.6%) in placebo-treated patients (P = 0.38). In the subgroup of 38 patients (37 pairs of evaluable ECGs) presenting more than 4h after symptom onset, ST-segment resolution was more pronounced in the patients randomized to intracoronary adenosine [median 73.8% (IQR, 55.9–79.5%)] than in the patients randomized to intracoronary placebo [median 32.6% (IQR, 15.3–53.8%), P = 0.008]. This resulted in a higher proportion of patients with >70% ST-segment resolution in those randomized to adenosine [11/19 (57.9%) vs. 2/18 (11.1%) for the placebo group, P = 0.003]. In contrast, in patients presenting within 4h of symptom onset, median ST-segment resolution tended to be lower in those randomized to adenosine when compared with the placebo group [63.6% (IQR, 31.3–77.4%) vs. 73.2% (IQR, 54.7–86.6%), P = 0.32], resulting in a proportion of patients with >70% ST-segment resolution of 13/34 (38.2%) patients randomized to adenosine and 19/34 (55.9%) patients randomized to placebo (P = 0.14).

Figure 5

Myocardial reperfusion as assessed by ST-segment resolution.

Angiographic results

After primary PCI, TIMI flow grade 3 was present in 45 of the 56 patients (80.4%) randomized to intracoronary adenosine and in 47 of the 54 patients (87.0%) randomized to intracoronary placebo (P = 0.65; Table 4). In stratum 1 (<4 h), significantly more patients in the placebo group had TIMI flow grade 3 than in the adenosine group [33/36 (91.7%) vs. 27/36 (75.0%)] (P = 0.04). In stratum 2 (>4 h), 14 of the 18 patients in the placebo group had TIMI flow grade 3, when compared with 18 of the 20 patients in the adenosine group (P = 0.16). Median TIMI frame count was 30.3 (IQR, 19.0–53.0) and 28.0 (IQR, 18.5–45.0) (P = 0.52) for the adenosine- and placebo-treated patients, respectively. A myocardial blush grade 3 was observed after primary PCI in 18 of 52 (34.6%) patients treated with adenosine and in 21 of 51 (41.2%) patients treated with placebo (P = 0.98) (Figure 6). Myocardial blush grade 0 or 1 was present in 19 of 52 (36.5%) patients and 21 of 51 patients (41.2%), respectively (P = 0.98). This absence of significant differences in myocardial blush grade was independent of time from symptom onset.

View this table:
Table 4

Angiographic findings after primary percutaneous coronary intervention

AdenosinePlaceboP-value
TIMI flow grade0.65
00/56 (0.0%)1/54 (1.9%)
12/56 (3.6%)1/54 (1.9%)
29/56 (16.1%)5/54 (9.3%)
345/56 (80.4%)47/54 (87.0%)
0/12/56 (3.6%)2/54 (3.7%)0.96
2/354/56 (96.4%)52/54 (96.3%)
TIMI frame count(n = 47)(n = 44)0.52
Median (IQR)30.3 (19.0, 53.0)28.0 (18.5, 45.0)
Figure 6

Myocardial reperfusion as assessed by myocardial blush grade.

Biomarkers

Calculation of the area under the curve of cardiac markers as an estimation of infarct size was not performed if either the first or last measurement was missing. Other missing values were imputed by linear interpolation. The area under the curve for CK and CK-MB could be calculated for only 98 and for troponin I for 96 of the 110 patients and showed no differences between treatment groups (Table 5), irrespective of time from symptom onset.

View this table:
Table 5

Infarct size based on cardiac markers [median (interquartile range)]

AdenosinePlaceboP-value
CK maximum (IU/L)2170 (1043; 3986) (n = 56)1878 (1234; 2967) (n = 54)0.24
CK-MB maximum (µg/L)244 (133; 441) (n = 56)227 (141; 405) (n = 54)0.47
Troponin I maximum (µg/L)91 (36; 176) (n = 56)78 (44; 112) (n = 54)0.22
CK-AUC (IU/L × h)43613 (19724; 68167) (n = 47)33220 (19695; 54298) (n = 51)0.25
CK-MB AUC (µg/L × h)3607 (2402;7194) (n = 47)3524 (1800; 6336) (n = 51)0.47
Troponin AUC (µg/L × h)1486 (753; 2696) (n = 45)1272 (666; 1959) (n = 51)0.47
  • AUC, area under the curve.

Complications during primary percutaneous coronary intervention

An increase of chest pain during or immediately after selective administration of adenosine 4 mg distally to the occlusion site was not observed more frequently when compared with the intracoronary administration of placebo [5/56 (8.9%) vs. 5/54 (9.3%) (P = 0.97) (Table 6)]. Occlusion of a significant side branch, no reflow phenomenon, bradycardia, ventricular tachycardia, and ventricular fibrillation was seen in comparable numbers of patients in the two treatment groups. Distal embolization was numerically more frequent in patients randomized to intracoronary adenosine when compared with patients randomized to intracoronary placebo, the difference, however, not reaching statistical significance [6/56 (10.7%) vs. 1/54 (1.9%), P = 0.06]. The incidence of second- and third-degree atrioventricular block was also numerically higher in adenosine-treated patients [3/56 (5.4%) and 10/56 (17.9%)] than in placebo-treated patients [1/54 (1.9%) and 4/54 (7.4%)]. These conduction blocks disappeared within 2 min, without clinical sequelae.

View this table:
Table 6

Complications during primary percutaneous coronary intervention [n (%)]

Adenosine (n = 56)Placebo (n = 54)P-value
Increase of chest pain during study drug administration5 (8.9%)5 (9.3%)0.97
Occlusion of side branch ≥2 mm1 (1.7%)1 (1.9%)0.99
Distal embolization6 (10.7%)1 (1.9%)0.06
No reflow2 (3.6%)2 (3.7%)1.00
Bradycardia5 (8.9%)7 (13.0%)0.53
Second degree AV block3 (5.4%)1 (1.9%)0.33
Third degree AV block10 (17.9%)4 (7.4%)0.10
VT1 (1.8%)1 (1.8%)1.00
VF1 (1.8%)2 (3.7%)0.56
  • AV, atrioventricular; VT, ventricular tachycardia; VF, ventricular fibrillation.

Clinical outcome

Two patients died during initial hospitalization: one patient randomized to intracoronary adenosine developed electromechanical dissociation 12 h after stenting of the mid-left anterior descending (LAD), complicated by no-reflow phenomenon. Abciximab had been discontinued early because of a frontal epidural hematoma in the medical history. One patient randomized to intracoronary placebo died of multi-organ failure after 7 days as a consequence of protracted cardiogenic shock.

Between discharge and 30-day follow-up, an additional patient randomized to intracoronary placebo died of a recurrent myocardial infarction 9 days after the index infarction.

At 1 year, two patients had died in each treatment group, without any apparent differences in functional status between surviving patients (Table 7).

View this table:
Table 7

Clinical outcome

AdenosinePlaceboP-value
At 30 days
 Mortality1/56 (1.8%)2/54 (3.7%)0.32
NYHA class
 I46/55 (83.6%)43/52 (82.7%)0.56
 II9/55 (16.4%)7/52 (13.5%)
 III0/55 (0%)2/52 (3.9%)
At 1 year
 Mortality2/56 (3.6%)2/54 (3.7%)0.97
NYHA class
 I48/54 (88.9%)47/52 (90.4%)0.71
 II5/54 (9.3%)4/52 (7.7%)
 III–IV1/54 (1.9%)1/52 (1.9%)
  • NYHA, New York Heart Association.

Discussion

This prospective, randomized, double-blind, placebo-controlled clinical trial demonstrated that high-dose (4 mg) selective intracoronary administration of adenosine distally to the occlusion site and immediately before initial balloon inflation did not enhance myocardial salvage in patients with STEMI. There were no significant differences in MSI, MVO, and infarct size as assessed by MRI, nor in ST-segment elevation resolution, TIMI flow and myocardial blush grade, enzymatic infarct size, and clinical outcome at 30 days and 1 year after primary PCI between patients randomized to intracoronary adenosine or intracoronary placebo. On the contrary, a small non-significant trend favouring patients randomized to placebo was observed for most secondary endpoints.

Adenosine and adenosine analogues have been shown in animal models to mimic ischaemic preconditioning and to potentially prevent reperfusion injury.8,9,2326 Lethal myocardial reperfusion injury is defined as myocardial injury caused by the restoration of coronary blood flow after an ischaemic episode, and is thought to account for up to 50% of the final size of a myocardial infarct. While the pathophysiology of this phenomenon is incompletely understood, an important role has been attributed to the reperfusion injury salvage kinase (RISK) pathway and the mitochondrial permeability transition pore (MPTP).27 These have become new targets for cardioprotection, and inhibition of the opening of the MPTP by means of cyclosporine seems to limit infarct size in humans.28 In two out of the three different proposed cardioprotective protein kinase pathways, adenosine is postulated to directly and indirectly activate protein kinase C, or to contribute to the activation of the RISK pathway.29 In addition, adenosine is a strong coronary vasodilator, and inhibits pre-synaptic release of norepinephrine from sympathetic nerves, platelet aggregation, and leucocyte adherence to the vascular wall.30

Clinical studies with adenosine in humans with STEMI have yielded mixed results. In the AMISTAD-II trial, conducted in patients with anterior STEMI treated by fibrinolysis or primary PCI, a 3 h intravenous infusion of adenosine did not reduce the primary clinical endpoint.10 However, in a post hoc subgroup analysis, a benefit was seen in those reperfused within 3 h, and a significant reduction in infarct size was reported in the high-dose adenosine subgroup.11 Several retrospective and prospective clinical studies with intracoronary administration of adenosine during primary PCI have shown favourable effects, such as increases in coronary blood flow, a reduction of no-reflow, improvement of signs of reperfusion on the ECG after PCI, and a reduction in infarct size.1315 However, these favourable results should be interpreted with caution because they are limited by a small sample size, potential selection bias, or a non-randomized trial design. A recent single-blind study of 448 STEMI patients randomized to two bolus injections of intracoronary adenosine (2 × 120 µg) or matching placebo after manual thrombus aspiration did not show any difference in the primary endpoint of residual ST-segment elevation after primary PCI.31 In addition, all secondary endpoints (ST-segment elevation resolution, myocardial blush grade, TIMI flow after PCI, enzymatic infarct size, and clinical outcome at 30 days) did not differ between treatment groups. As a possible explanation, the authors argued that adenosine may have a maximal beneficial effect on ischaemic myocardial injury when given before reperfusion therapy, whereas they performed thrombus aspiration before administration of adenosine or placebo. Different methodologies to assess infarct size could also explain variable results in clinical studies. Magnetic resonance imaging is superior for determining spatial extent and transmural distribution of infarction over previously used techniques, including radionuclide imaging, wall motion at rest, enzymatic infarct size, and ST-segment resolution. In fact, MRI is the most accurate non-invasive technique to measure infarct size relative to risk area in humans.1618 This measurement is a robust and clinically relevant indicator of therapeutic benefit, and is expressed as MSI. For example, in a patient with a very large AAR, final infarct size can be substantial despite significant salvage. Conversely, in a patient with a very small AAR, infarct size will always be very small, irrespective of degree of salvage.

In the present study, we attempted to avoid some of the methodological limitations of earlier trials. First, we used MRI to assess the primary endpoint, myocardial salvage, because it is less influenced by differences in the extent of ischaemic myocardium at baseline. Second, all study personnel remained blinded to treatment assignment until completion of all analyses. Third, we administered the study drug subselectively to the ischaemic myocardium, prohibiting inadvertent spill-over into other coronary branches. Fourth, we used a very high dose of adenosine, 4 mg, proven to be safe and well tolerated.13 And last, we administered the intracoronary adenosine before initial balloon inflation.

One may argue that the absence of a beneficial effect of intracoronary adenosine is due to differences in baseline clinical characteristics, favouring placebo-treated patients. However, there were no significant differences in baseline characteristics between the two treatment groups with the exception of a slightly higher heart rate on admission in patients randomized to adenosine and a higher incidence of diabetes mellitus and previous PCI in patients randomized to placebo. Also, we observed a greater incidence of LAD involvement in patients randomized to adenosine, but the possibility of larger AAR could not be substantiated by quantification of oedema volume on T2-weighted MRI images.1618

In subgroup analyses for MSI, we found a statistically significant interaction between treatment and baseline TIMI flow grade (0/1 vs. 2/3) (P = 0.0056), whereby a statistically significant treatment benefit of adenosine was found for patients with TIMI 2/3 flow at baseline (P = 0.0486). While this finding concerns a small group of patients and may be due to multiple testing without correction of P-values, we cannot exclude the possibility of a true beneficial effect of adenosine in patients with spontaneous reperfusion, and of harm in patients without spontaneous reperfusion. One may hypothesize that an intact microcirculation before adenosine infusion allowing its distal delivery is important for an eventual beneficial effect to be seen. This would make it more difficult, however, to understand the positive outcome with intracoronary adenosine in a smaller but very similar study, in which patients with TIMI 3 flow at baseline were excluded.13

We observed significantly better ST-segment resolution in adenosine-treated patients presenting more than 4 h after symptom onset. However, in these patients, there was not even a trend of improved coronary blood flow, more myocardial salvage, decreased MVO, or smaller enzymatic infarct size. Therefore, and because there are no pathophysiological grounds to expect adenosine to be more beneficial in STEMI patients who present later, we assume that this observed difference is due to chance.

A possible explanation for the observed lack of beneficial effect of intracoronary adenosine may be that high-level endogenous production of adenosine during myocardial ischaemia already mediated maximal vasodilatation. Animal models demonstrate that adenosine concentrations may increase to levels producing maximal coronary arteriolar dilatation during ischaemia.32 Therefore, our study may be neutral because of the absence of vasodilatory reserve during and after reperfusion by PCI.

It is unlikely that the dose of adenosine in this study was insufficient to induce maximal vasodilatation because it is 40 times higher than the >100 µg intracoronary adenosine doses routinely recommended for fractional flow reserve measurements.33 The half-life of adenosine in human blood, 1 s, is apparently too short to protect from oxygen free radicals, which peak at 2–3 min of reperfusion, even if reperfusion is started only 15–30 s after adenosine administration. However, half-life may underestimate the duration of the biological effects of adenosine. After a bolus injection of adenosine in a non-stenotic coronary artery, coronary flow remains elevated for several minutes, suggesting persistent receptor activation well beyond half-life. In case of injection into a stenotic coronary artery, flow remains elevated even longer. In this study, adenosine was injected into a vascular bed with minimal, if any, antegrade flow, and under these circumstances, its biological effects could last much longer than expected on the basis of its half-life. The short half-life of adenosine, however, provides an intrinsic protection from possible side effects. In addition, ischaemic pre- and post-conditioning share a common signalling pathway that includes—among others—adenosine receptors. This signalling is thought to protect by suppressing the opening of MPTPs at reperfusion.

Alternatively, timing of adenosine administration may affect outcome. Previous studies have suggested that the window of opportunity is limited to the first 3–4 h after onset of ischaemic symptoms.11,34 Although most of our patients (65.5%) were treated within this time window, we failed to document benefit, even in a separate analysis of patients who underwent primary PCI within 4 h from symptom onset.

Limitations of the study

First, the primary endpoints were myocardial salvage and MVO and we cannot comment on clinical endpoints, for which the study was underpowered. Second, 28% of patients had limited spontaneous reperfusion before intracoronary administration of the study drug and in a few more mere passage of the guide wire may have mediated some degree of reperfusion. Hence, in these patients, adenosine was not truly given before reperfusion. However, for MSI, we identified a large interaction between treatment and baseline TIMI, whereby a statistically significant treatment benefit of adenosine was found for patients with TIMI 2/3 flow at baseline and an almost statistically significant worsening effect of adenosine in the TIMI 0/1 group. Although it is not clear to us how this finding can be explained, it makes it very unlikely that some degree of reperfusion before the administration of adenosine abolished an eventual beneficial effect. Third, the trial included a significant proportion of patients with small and/or completed infarcts. Fourth, infarct size cannot be reliably assessed when biomarkers are drawn every 8 h as in this study. And finally, adenosine boluses may be too short acting for any beneficial effect and we cannot exclude efficacy of prolonged intracoronary administration.

Conclusions and implications

We found no evidence that selective intracoronary administration of high-dose adenosine as adjunctive therapy to primary PCI enhances myocardial salvage or reduces MVO in patients with STEMI. Future attempts at improving myocardial reperfusion, preventing reperfusion injury, and salvaging ischaemic myocardium should focus on pharmacological therapies with other modes of action or non-pharmacological approaches.

Conflict of interest: none declared.

References

View Abstract