OUP user menu

★ editor's choice ★

Pretreatment with intracoronary adenosine reduces the incidence of myonecrosis after non-urgent percutaneous coronary intervention: a prospective randomized study

Chi-Hang Lee, Adrian Low, Bee-Choo Tai, Melissa Co, Mark Y. Chan, Jimmy Lim, Yean-Teng Lim, Huay-Cheem Tan
DOI: http://dx.doi.org/10.1093/eurheartj/ehl411 19-25 First published online: 28 November 2006


Aims We sought to investigate the effect of adenosine pretreatment on the incidence of myonecrosis after non-urgent percutaneous coronary intervention (PCI).

Methods and results This was a prospective, randomized, open-label study. Patients who were scheduled for non-urgent PCI in de novo native coronary arteries were eligible. All patients were pretreated with aspirin and clopidogrel. Myonecrosis was measured by creatine kinase-myocardial band (CK-MB) elevation after PCI. A total of 62 patients were randomized into the adenosine (n = 31) or standard (n = 31) group. The adenosine group received 50 µg adenosine bolus before wiring of each lesion, whereas the standard group did not. Post-PCI myonecrosis occurred more frequently in the standard group (39 vs. 13%, OR 0.23, 95% CI 0.05–0.95, P = 0.020). After adjustment for drug-eluting stent implantation, multi-vessel stenting, and elevated baseline troponin, the OR was 0.19 (95% CI 0.05–0.72, P = 0.017). The median peak values of CK-MB in the adenosine and standard groups were 2 and 4 µg/L, respectively (P = 0.033). The adjusted difference was 1.95 µg/L (95% CI 0.13–3.77, P = 0.037). The incidences of myocardial infarction (>3 × CK-MB) were 6 and 16% in the adenosine and standard groups, respectively (OR 0.36; 95% CI 0.03–2.46, P = 0.229).

Conclusion Pretreatment with 50 µg of adenosine decreases the incidence of myonecrosis after non-urgent PCI compared with that without pretreatment.

  • Adenosine
  • Myonecrosis
  • Myocardial infarction
  • Coronary intervention


Percutaneous coronary intervention (PCI) is associated with up to 30% incidence of myonecrosis, as reflected by elevation of cardiac enzymes in an otherwise successful procedure.1 Apart from side-branch occlusion, intimal dissection and coronary spasm, a possible aetiology of myonecrosis after PCI might be distal embolization of atherogenic materials from plaque disruption,2 causing obstruction of blood flow at capillary level resulting in micro-infarction.3,4

Recent studies have suggested that pretreatment with adenosine may be associated with a reduction in infarct size after reperfusion therapy for acute myocardial infarction.58 Whether pretreatment with adenosine decreases the incidence of myonecrosis in patients undergoing non-urgent PCI is not fully known.9 We propose that adenosine-induced hyperaemia can potentially ameliorate the deleterious effects of distal embolization associated with non-urgent PCI through dilatation of the microvasculature. Mechanistically, this may reduce capillary obstruction by facilitating the throughput passage of embolized platelet thrombi out to the venous end of the coronary circulation, thereby reducing the incidence of post-PCI myonecrosis.

In this prospective, randomized, open-label study, we evaluated the incidence of myonecrosis after non-urgent PCI with and without pretreatment with intracoronary adenosine.


Eligibility criteria

Patients (aged between 21 and 75 years) who were scheduled to undergo non-urgent PCI of de novo native coronary lesions for stable or unstable angina were considered for recruitment into the study. Inclusion criteria included the following: baseline creatine kinase (CK) and creatine kinase-myocardial band (CK-MB) had to be within normal limits (a normal CK and CK-MB and elevated troponin allowed inclusion). We excluded patients with complete occlusion resulting in Thrombolysis In Myocardial Infarction (TIMI) grade 0 antegrade flow, thrombus-laden lesions, significant left main coronary stenosis, left ventricular ejection fraction < 30%, inability to give informed consent, bradycardia with heart rate below 50 b.p.m., allergy to adenosine, and the occurrence of myocardial infarction within one week. The institutional review board of the local institution approved the study and all patients gave informed consent before the procedure.

Study design

On the basis of a two-sided test size of 5% and a power of 80%, it was anticipated that a minimum of 62 patients would need to be recruited in each group to detect a difference in the incidence of myonecrosis of 20%. This assumed that the incidence of myonecrosis was 30% in the standard group, and 10% in the adenosine group. Eligible patients were randomly allocated to receive adenosine or standard treatment using block allocation. The treatment was assigned based on a 1:1 ratio. Applying the Pocock group sequential design,10 for a trial with two planned analyses (the first after half the number of patients were recruited) would assume P < 0.029 as a stopping rule at each analysis for a treatment difference.


Patients randomized to the adenosine group were given 50 µg of adenosine (diluted into 5 mL normal saline) through the guiding catheter into the target coronary artery prior to guidewire advancement. For multi-vessel stenting, another 50 µg of adenosine would be given prior to intervention of the subsequent vessel. The rest of the intervention strategy, including the use of glycoprotein IIb/IIIa inhibitors and drug-eluting stents, was left to the discretion of individual operators. Patients randomized to the standard therapy did not receive adenosine pretreatment, and the procedure was carried out in the usual manner. Administration of adenosine for treatment of no-reflow phenomenon during the procedure was allowed in both groups. All patients received a clopidogrel loading dose of 300 mg before the PCI (for loading dose administered more than 6 h prior to procedure),11 or 600 mg clopidogrel (for loading dose less than 6 h before procedure).12 Post-procedural antiplatelet regimen consisted of aspirin at 100 mg/day indefinitely and clopidogrel 75 mg/day for at least one month.


Following the procedure, blood samples for CK and CK-MB were collected between 12 to 24 h post PCI. CK-MB values were considered abnormal if they were elevated above the upper limit of normal (ULN). This is set at 6 µg/L by our local laboratory. If the first blood sample showed a CK-MB level ≥ 12 µg/L (≥2 times upper normal limit), a second blood sample would be drawn every 8 h later until a downward trend was observed. For patients with two or more blood samples drawn, the peak CK-MB level was used for analysis.

The primary objective of this study was to evaluate the incidence of myonecrosis (any elevation of CK-MB above 6 µg/L), in patients undergoing non-urgent PCI with or without pretreatment with intracoronary adenosine. Secondary endpoints included median peak CK-MB level and post-procedural myocardial infarction. The latter is defined by post-procedural increase of CK-MB more than three times above the ULN. Adverse events including death of any cause, myocardial infarction, unplanned target-lesion revascularization within 30 days of intervention were determined by telephone interview and/or medical record review.

Statistical methods

The incidence of myonecrosis in the adenosine and standard groups were compared using odds ratio (OR), and its associated 95% confidence interval (CI). Logistic regression analysis was used to account for the effects of possible confounders such as multi-vessel stenting, implantation of drug-eluting stent, stent length, and raised troponin in patients with unstable angina. These parameters were chosen a priori based on clinical grounds. Multi-vessel stenting and unstable angina are well believed to be associated with a higher incidence of post-procedural myonecrosis. There is recent evidence suggesting that drug-eluting stents might be associated with increased risk of post-procedural myocardial infarction.13 Of the variables under consideration, multi-vessel stenting showed the strongest association with myonecrosis, and stent length the weakest. However, including all these variables in the model resulted in unexpectedly large standard errors, indicating a problem with collinearity between stent length and multi-vessel stenting. To resolve this problem, we excluded stent length (the weakest of all variables) from the model. Median peak CK-MB measurements were compared between the two treatment regimens, and quantile regression was used to adjust for the effects of multi-vessel stenting, implantation of drug-eluting stents, raised troponin in patients with unstable angina, and the number of CK-MB measurements taken to achieve the peak value. Statistical significance was evaluated at the conventional 5% level based on a two-sided test. All statistical analyses were generated based on intention-to-treat using STATA version 8.2.


A total of 62 patients (adenosine group, n = 31; standard group, n = 31) were recruited into this study between 1 January and 31 March, 2006 (Figure 1). The data of these 62 patients were reviewed and analysed as defined a priori. All patients complied fully with the allocated treatment strategies and subsequent evaluations. The baseline clinical, angiographic, and procedural characteristics of the patients were similar in the two groups (Tables 1 and 2). Seven patients in each group received 600 mg clopidogrel (<6 h before PCI) and the rest received 300 mg clopidogrel (>6 h before PCI) loading. All patients were on statins prior to PCI.

View this table:
Table 1

Demographic and clinical characteristics of patients by treatment

Adenosine (n = 31)Standard (n = 31)P-value
Mean age in years (SD)56.9 (9.3)61.1 (10.9)0.114
Male sex (n, %)24 (77)24 (77)1.000
Smoking (n, %)17 (55)12 (39)0.203
Hypertension (n, %)21 (68)23 (74)0.576
Diabetes mellitus (n, %)15 (48)19 (61)0.307
Hypercholesterolaemia (n, %)27 (87)27 (87)1.000
Family history of CAD(n, %)0 (0)0 (0)
Indication (n, %)0.864
 Stable angina22 (71)20 (65)
 Unstable angina9 (29)11 (35)
  Positive troponin T2 (6)1 (3)
  Negative troponin T7 (23)10 (32)
Previous myocardial infarction (n, %)1 (3)3 (10)0.301
Clopidogrel pre-treatment regimen (n, %)1.000
 300 mg > 6 h before procedure24 (77)24 (77)
 600 mg < 6 h before procedure7 (23)7 (23)
View this table:
Table 2

Baseline angiographic and procedural characteristics of patients by treatment

Adenosine (n = 31)Standard (n = 31)P-value
Site of puncture (n, %)0.199
 Femoral27 (87)23 (74)
 Radial4 (13)8 (26)
ACC/AHA lesion classification (n, %)0.192
 A3 (10)0 (0)
 B17 (23)4 (13)
 B212 (39)17 (55)
 C9 (29)10 (32)
Multi-vessel stenting (n, %)9 (29)10 (32)0.783
Drug-eluting stent use (n, %)12 (39)15 (48)0.442
Number of stent implanted (n, %)0.420
 121 (68)16 (52)
 29 (29)13 (42)
 31 (3)2 (6)
Side-branch compromised, TIMI flow <3 (n, %)4 (13)3 (10)0.688
Baseline TIMI flow (n, %)0.717
 11 (3)2 (6)
 22 (6)1 (3)
 328 (90)28 (90)
Final TIMI flow (n, %)-
 331 (100)31 (100)
Post-stenting balloon dilatation (n, %)11 (35)5 (16)0.082
Ostial lesion (n, %)1 (3)3 (10)0.301
Angulated lesion, >45 degree (n, %)3 (10)2 (6)0.641
Calcified lesion (n, %)5 (16)4 (13)0.718
Bifurcation lesion (n, %)8 (26)11 (35)0.409
Proximal tortuosity (n, %)0 (0)2 (6)0.151
Median stent length (mm) (range)23 (12–84)33 (12–76)0.150
Median fluoroscopy time (second) (range)1166 (319–5200)1402 (530–4420)0.492
Median contrast amount (mL) (range)140 (60–370)165 (80–280)0.329

After adenosine administration, no patient developed significant bradycardia or heart block requiring pacing. The distribution of the number of stents implanted was similar in the adenosine and standard groups. The median stent length was longer in the standard group (33 vs. 23 mm). Drug-eluting stents were implanted in 12 (39%) and 15 (48%) patients in the adenosine and standard groups, respectively. Multi-vessel stenting was performed in nine (29%) and 10 (32%) patients in the adenosine and standard groups, respectively. None of the patients in this study received glycoprotein IIb/IIIa inhibitors. All procedures were successful and uncomplicated. Slow-flow or no-reflow phenomenon did not occur in all procedures in this study. Therefore, bail-out nitroprusside, verapamil, or adenosine was not used in any of the procedures. No patient developed contrast nephropathy. Quantitative coronary angiography data are shown in Table 3.

View this table:
Table 3

Quantitative coronary angiography

Baseline reference vessel diameter, mm2.90 (0.53)2.42 (0.66)0.003
Baseline percent stenosis, %85.7 (9.18)83.0 (9.12)0.263
Baseline minimal lumen diameter, mm0.42 (0.31)0.41 (0.26)0.900
Final reference vessel diameter, mm3.09 (0.51)2.82 (0.61)0.069
Final percent stenosis, %4.38 (3.04)6.67 (7.62)0.135
Final minimal lumen diameter, mm2.95 (0.51)2.64 (0.64)0.040
Acute gain, mm2.53 (0.53)2.23 (0.61)0.043
  • Results are expressed as mean (SD).

In the standard arm, six (19%) patients had CK-MB elevation > 2 times the ULN and therefore had the second sample for CK-MB drawn. In all six cases, the second sample already showed a downward trend for the CK-MB level. In the adenosine arm, two (6%) patients had CK-MB elevation more than two times the ULN and therefore had the second sample for CK-MB drawn. Similarly, in all two cases the second sample already showed a downward trend for the CK-MB level.

Post-PCI myonecrosis occurred more frequently in the standard (39%, n = 12) than in the adenosine group (13%, n = 4) (Figure 2). The OR for myonecrosis comparing adenosine with standard treatment was 0.23 (95% CI 0.05–0.95, P = 0.020). Taking into account possible confounding effects of drug-eluting stent implantation, multi-vessel stenting, and elevated troponin levels in patients with unstable angina, the adjusted OR was 0.19 (95% CI 0.05–0.75, P = 0.017) (Table 4). The median peak values of CK-MB in the adenosine and standard groups were 2 and 4 µg/L, respectively (P = 0.033), with a notably larger spread of values for the standard group (Figure 3). The adjusted difference was 1.3 U/L (95% CI 0.44–2.16, P = 0.004). The incidences of myocardial infarction (>3 × CK-MB) were 6% (n = 2) and 16% (n = 5) in the adenosine and standard groups, respectively (OR 0.36; 95% CI 0.03–2.46, P = 0.229). The results remained unaltered after adjusting for these covariates: drug-eluting stent implantation, multi-vessel stenting, and raised troponin in patients with unstable angina.

Figure 2

Incidences of CK and CK-MB elevation after non-urgent PCI in the adenosine and standard groups.

Figure 3

Distribution of CK-MB level after PCI in patients in the adenosine and standard groups.

View this table:
Table 4

OR estimates of myonecrosis for the effect of adenosine treatment, multi-vessel stenting, drug-eluting stent implantation, and indication of angina

Myonecrosis N (%)OR (95% CI)
Crude estimateAdjusted estimate
 Adenosine4 (13)0.23 (0.05–0.95)0.19 (0.05–0.75)
 Standard12 (39)1.001.00
Multi-vessel stenting
 Yes3 (16)0.43 (0.11–1.74)0.35 (0.07–1.70)
 No13 (30)1.001.00
Implantation of drug-eluting stent
 Yes5 (19)0.50 (0.15–1.65)0.41 (0.11–1.53)
 No11 (31)1.001.00
 Unstable angina
  Positive troponin T1 (33)1.60 (0.13–19.56)1.55 (0.09–26.39)
  Negative troponin T5 (29)1.33 (0.38–4.71)1.53 (0.35–6.67)
 Stable angina10 (24)1.001.00

Apart from patients with documented post-procedural myocardial infarction (>3 × CK-MB), there were no deaths, myocardial infarction, or recurrent ischaemia requiring urgent target lesion revascularization during both the in-hospital period and at 30-day follow-up.


This is the first randomized trial investigating the impact of adenosine bolus pretreatment on the incidence of myonecrosis after non-urgent PCI. We found that pretreatment with 50 µg of adenosine decreases the incidence of myonecrosis after PCI to de novo native coronary arteries, compared with standard PCI without pretreatment. No major adverse effects related to adenosine were observed.

PCI has been established as an effective treatment strategy for obstructive coronary artery disease. However, PCI is associated with an up to 30% incidence of myonecrosis, as reflected by the elevation of post-procedural CK-MB.1 Aetiologies of myonecrosis after PCI are likely to be multi-factorial and remain incompletely understood. Apart from side-branch occlusion, intimal dissection, and coronary spasm, studies using contrast-enhanced magnetic resonance imaging suggest that distal embolization of atherogenic and/or thrombogenic materials resulting in micro-infarction might be an important mechanism.3,4 Myonecrosis of varying extents has been shown to be associated with an increased risk of future adverse cardiac events including death and myocardial infarction.1418

Currently, pretreatment with antiplatelet agents and statins are strategies proven to reduce the incidence of post-PCI myonecrosis.1921 In a randomized trial, loading with 600 mg of clopidogrel before PCI was associated with a significant reduction in the incidence of myonecrosis from 26 to 14%, compared with 300 mg loading.19 Similarly, pretreatment with 40 mg of atorvastatin for 7 days prior to PCI was shown to reduce the myonecrosis risk from 35 to 12%.20 To our knowledge, this is the first study documenting an incremental benefit of adenosine in further reducing myonecrosis in patients pretreated with antiplatelet agents and statins.

Adenosine is a naturally occurring nucleoside with a half-life in blood of less than 10 s. Adenosine, administered via the intravenous or intracoronary route, produces a hyperaemic effect that is commonly used in the measurement of coronary flow reserve and fractional flow reserve during PCI. Recent studies suggested that adenosine infusions might be associated with a reduction in heart muscle damage following reperfusion therapy for acute myocardial infarction.58 Some of these studies however have not demonstrated a significant benefit of adenosine pretreatment. We believe the failure of some of these previous studies to show a positive effect is due to the overwhelming amount of red thrombi that embolizes during acute myocardial infarction. This thus negates any beneficial effect that adenosine-induced hyperaemia may potentially have. More importantly, in many of these studies, adenosine was only administered after the onset of acute infarction, by which time significant distal embolization would have already occurred.

The possibility that pretreatment with adenosine decreases myonecrosis in patients undergoing non-urgent PCI has not been properly studied so far. One animal study suggested endogenous release of adenosine could compensate for the distal embolization of small particles.22 With adenosine pretreatment before guidewire advancement, we believe the hyperaemic effect might be sufficient to ameliorate distal embolization of relatively small platelet thrombi following non-urgent PCI. A small study involving 28 patients has previously suggested that adenosine infusion might be able to reduce the CK-MB elevation after non-urgent PCI.9 Instead of using continuous intravenous infusion, we used bolus doses of adenosine at 50 µg, administered through the guiding catheter before PCI to each lesion. Although it is conceivable that only a fraction of the administered adenosine would reach the target distal vessel, this dosage and route of administration has been found to be sufficient to achieve maximum hyperaemia.23 In a study involving 191 patients and 215 vessels, the minimum dose of adenosine required to achieve minimum fractional flow reserve (or maximum hyperaemia) was 42 µg in both left and right coronary arteries.23 Moreover, an intracoronary bolus approach has advantages of ease of administration, lower cost, and reduced risk of bradycardia. Our results showed that the incidence of myonecrosis was reduced from 39 to 13% with adenosine pretreatment. Finally, the procedural time as reflected by fluoroscopy time and amount of contrast used was not increased in the adenosine group.

The mechanism of the protective effect of adenosine remains elusive. Postulated mechanisms include hyperaemia,24 preconditioning,25 anti-platelet effect,26 and anti-inflammatory effect.27 There is evidence from previous studies that despite a short serum half-life, the cardioprotective effect of adenosine persists much longer. Leesar et al.25 explored the ischaemic preconditioning effect of intracoronary adenosine in human subjects. Beneficial effects were observed in patients undergoing balloon angioplasty when adenosine was administered prior to the procedure. Similarly, in the ADELINE pilot trial where pre-treatment with adenosine was shown to reduce post-procedural CK-MB elevation, with adenosine administered 10 min before the first balloon inflation.9

The incidence of myonecrosis in the control group (39%) in the present study was relatively high compared with another study that had similar inclusion and exclusion criteria.19 We believe this is because the PCI procedures in our study were more complex, as evidenced by more multi-vessel PCI, a higher number of stents implanted, and the use of longer stents.


Our study has several limitations. First, the study population was small and the design was an open-label study. However, despite the small sample size, we managed to show a significant reduction in the primary endpoint, assuming a stopping rule of P < 0.029. Further recruitment of subjects was not initiated due to ethical considerations. In addition, we have also taken the necessary precautions to adjust for any potential confounding effects via multivariable analysis to account for other pertinent factors. During the design of the study, the authors did consider the option of double-blind study. Apart from the cost issue, we concluded that it was almost impossible to blind the operators as most patients would develop transient heart block and bradycardia after adenosine administration. The arrhythmic effect could be easily observed on the monitoring screen where the operators have to pay close attention to during the coronary intervention procedure. A second limitation is that our study population was Asian in origin, whereas all other studies on myonecrosis included predominantly Caucasian patients. The results of our study may thus not be applicable to a broader range of ethnic groups. Third, inflammatory markers, such as C-reactive protein, were not assessed in this study. However, the mechanism of action of adenosine is believed to be hyperaemia and not anti-inflammatory. Besides, all study patients were on statins therapy prior to PCI and this is well documented to reduce inflammation. Fourth, none of the study patients received glycoprotein IIb/IIIa inhibitors, reflecting current practice patterns at our centre following recent studies showing an absence of additional benefit with IIb/IIIa inhibitors in patients adequately treated with antiplatalet therapy before non-urgent PCI.2830 Finally, post-PCI troponin level was not measured, but its prognostic significance in this setting remains controversial.3133


Bolus administration of 50 µg of adenosine through the guiding catheter is associated with a 77% reduction in the occurrence of myonecrosis following non-urgent PCI in patients pretreated with dual antiplatelet agents and statins. Further studies are required to investigate the protective effect of adenosine in patients not pretreated with antiplatelet agents including those intolerant or allergic to aspirin or clopidogrel, and those with aspirin and/or clopidogrel resistance.

Conflict of interest: none declared.


  • The abstract of this study was presented during the Transcatheter Cardiovascular Therapeutics 2006, October 22–27, at Washington Convention Center, Washington DC, USA.


View Abstract