Adenosine and verapamil for no‐reflow during primary percutaneous coronary intervention in people with acute myocardial infarction

Abstract

Background

Primary percutaneous coronary intervention (PPCI) is the preferred treatment for ST‐segment elevation myocardial infarction. Although coronary flow is restored after PPCI, impaired myocardial perfusion (known as no‐reflow) related to poor clinical outcomes is frequently observed. To overcome this phenomenon, drugs, such as atorvastatin, abciximab and others, have been tried as adjunctive treatment to PPCI. Among these drugs, verapamil and adenosine are among the most promising. No other systematic reviews have examined use of these two drugs in people with acute myocardial infarction (AMI) undergoing PPCI. This is an update of the version previously published (2013, Issue 6), for which the people of interest in the review were those treated with PPCI ‐ not those given fibrinolytic therapy.

Objectives

To study the impact of adenosine and verapamil on no‐reflow during PPCI in people with AMI.

Search methods

We updated searches of the following databases in June 2014 without language restriction: the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, EMBASE, Web of Science and BIOSIS, China National Knowledge Infrastructure and clinical trials registers (ClinicalTrials.gov, Current Controlled Trials, Australian and New Zealand Clinical Trials Registry, the World Health Organization (WHO) International Clinical Trials Registry Platform). We also handsearched The American Journal of Cardiology.

Selection criteria

We selected randomised controlled trials (RCTs) in which adenosine or verapamil was the primary intervention. Participants were individuals diagnosed with AMI who were undergoing PPCI.

Data collection and analysis

Two review authors collected studies and extracted data. When necessary, we contacted trial authors to obtain relevant information. We calculated risk ratios (RRs), P values and 95% confidence intervals (CIs) of dichotomous data.

Main results

We included in our review 11 RCTs (one new study with 59 participants) involving 1027 participants. Ten RCTs were associated with adenosine and one with verapamil. We considered the overall risk of bias of included studies to be moderate. We found no evidence that adenosine reduced short‐term all‐cause mortality (RR 0.61, 95% CI 0.25 to 1.48, P value = 0.27), long‐term all‐cause mortality (RR 0.78, 95% CI 0.22 to 2.74, P value = 0.70), short‐term non‐fatal myocardial infarction (RR 1.32, 95% 0.33 to 5.29, P value = 0.69) or myocardial blush grade (MBG) 0 to 1 after PPCI (RR 0.96, 95% CI 0.76 to 1.22, P value = 0.75). The incidence of thrombolysis in myocardial infarction (TIMI) flow grade < 3 after PPCI (RR 0.62, 95% CI 0.42 to 0.91, P value = 0.01) was decreased. Conversely, adverse events with adenosine, such as bradycardia (RR 6.32, 95% CI 2.98 to 13.41, P value < 0.00001), hypotension (RR 11.43, 95% CI 2.75 to 47.57, P value = 0.0008) and atrioventricular (AV) block (RR 6.78, 95% CI 2.15 to 21.38, P value = 0.001), were significantly increased.

Meta‐analysis of verapamil as treatment for no‐reflow during PPCI was not performed because data were insufficient.

Authors' conclusions

It is difficult to draw conclusions because of the insufficient quality and quantity of current research studies. We considered the overall risk of bias of included studies to be moderate. Adenosine as treatment for no‐reflow during PPCI could reduce angiographic no‐reflow (TIMI flow grade < 3) but was found to increase adverse events. What's more, no evidence could be found to suggest that adenosine reduced all‐cause mortality, non‐fatal myocardial infarction or the incidence of myocardial blush grade 0 to 1. Additionally, the efficacy of verapamil for no‐reflow during PPCI could not be analysed because data were insufficient. Further clinical research into adenosine and verapamil is needed because of the limited numbers of available trials and participants.

Plain language summary

Drugs used as add‐on therapy to heart procedure following heart attack

Acute myocardial infarction (AMI), or heart attack, is one of the major causes of mortality worldwide. Approximately one‐third of people suffering heart attacks die before they reach the hospital. Primary percutaneous coronary intervention (PPCI) is a procedure whereby the coronary artery is widened without surgery, using a stent. Although this procedure restores blood flow through the coronary artery, perfusion through all areas of the heart may not occur. This is known as no‐reflow. To try to resolve this problem, healthcare workers have tried drugs such as adenosine and verapamil as add‐on treatment. We include 11 studies in this review with a total of 1027 participants. Ten studies compared adenosine or placebo as an addition to PPCI, and one compared verapamil or placebo.

No evidence suggests that adenosine reduced short‐term or long‐term all‐cause mortality or non‐fatal myocardial infarction to a greater extent than placebo treatment. However, we found evidence to show that adenosine can decrease angiographic no‐reflow after PPCI. Conversely, we also found evidence indicating that adverse events (bradycardia, hypotension, atrioventricular block) were increased with adenosine, although these effects were short‐lived. Included studies were few, with low numbers of participants and a relatively short period of follow‐up, and event rates of mortality and angiographic no‐reflow were low. Our meta‐analyses therefore had insufficient numbers to support a clear‐cut conclusion about the impact of adenosine in people with AMI undergoing PPCI.

Background

Description of the condition

Acute myocardial infarction (AMI), one of the major causes of mortality worldwide, can be defined from different perspectives related to clinical, electrocardiographic, cardiac biomarker and pathological characteristics. Different types of myocardial infarction have been identified: Type 1 is defined as spontaneous myocardial infarction related to ischaemia due to a primary coronary event such as plaque erosion and/or rupture, fissuring or dissection (Thygesen 2007). It is estimated that more than three million people have acute ST‐elevation myocardial infarction (STEMI) each year, and more than four million have non‐ST‐elevation myocardial infarction (White 2008). Approximately one‐third of people with myocardial infarction die before they reach the hospital and can receive effective treatment (Huikuri 2001). Rapid reperfusion of the infarcted coronary artery is required in STEMI and can be carried out by pharmacological or mechanical therapy; primary percutaneous coronary intervention (PPCI) is accepted as superior to fibrinolysis (Keeley 2003; Weaver 1997; Zijlstra 1993). People of interest for this review included those having PPCI ‐ not those given fibrinolytic therapy. PPCI is associated with normal epicardial flow in more than 90% of patients (Boersma 2006; Simes 1995). More than 50% of PPCI patients had door‐to‐balloon time (from time of arrival at the initial hospital to time of first balloon inflation at the PPCI hospital) of within 90 minutes (Rathore 2009), and less than 5% of interhospital‐transferred patients had undergone PPCI within 90 minutes after first medical contact (Boersma 2006). Although epicardial coronary flow can be restored after PPCI, impaired myocardial perfusion, known as the no‐reflow phenomenon, is frequently observed. The pathogenetic components of no‐reflow are very complicated and include a variable combination of distal atherothrombotic embolisation, ischaemic injury, reperfusion‐related injury and susceptibility of coronary microcirculation to injury (Niccoli 2009a). Diagnosis of this phenomenon can be confirmed by angiography, electrocardiography, myocardial contrast echocardiography and cardiac magnetic resonance (Niccoli 2009a). Prevalence ranges from 5% to 50% after PPCI, depending on the assessment methods used (Eeckhout 2001; Rezkalla 2008). In the present meta‐analysis, angiography was used to identify the no‐reflow phenomenon because data were available across all included studies. Angiographic no‐reflow can be defined as (1) a thrombolysis in myocardial infarction (TIMI) flow grade < 2 despite vessel patency and absence of dissection, spasm or distal macroembolus, or (2) a myocardial blush grade (MBG) of 0 to 1 (Niccoli 2009a). More important, AMI with no‐reflow is strongly associated with poor clinical outcomes (Henriques 2003; Morishima 1995).

Description of the intervention

To improve myocardial perfusion or to prevent no‐reflow after PPCI, many drugs, such as atorvastatin, tirofiban and others, have been tried clinically as adjunctive treatment to PPCI (Kunadian 2008). Both adenosine and verapamil have vasodilatory effects and have been commonly studied (Kunadian 2008).

Adenosine, an endogenous nucleoside, has vasodilatory actions on both arteries and arterioles, mediated through adenosine A_2A and A_2B receptors (Forman 2006). It reduces the release of vasoconstrictors by inhibiting activation of neutrophils and platelets, and it prevents endothelial damage and microvascular spasm, both of which contribute to the cardioprotective and antithrombotic effects of adenosine (Berne 1980; Minamino 1998). Adenosine can decrease free radical formation, thereby limiting the degree of myocardial injury (Forman 2006). Moreover, under some circumstances, it may act as an antiarrhythmic agent by blocking the atrioventricular (AV) nodes via adenosine A receptors, and it has inhibitory effects on the central nervous system (Mustafa 2009). Because of the extremely short half‐life of intravascular adenosine (one to two seconds in humans), its adverse effects are very short‐lived (Forman 2006). Adverse effects of intracoronary adenosine in people with AMI include hypotension (14.3%), bradycardia (20%), second‐degree AV block (18.5%) and bronchospasm (0.6%). Fortunately, these adverse effects last for only two to three minutes after adenosine is administered (Fokkema 2009).

Verapamil, a calcium channel blocker, can decrease oxygen demand by lowering heart rate and arterial pressure, and can improve myocardial perfusion by relieving microvascular spasm (Taniyama 1997), which may subsequently reduce infarct size. Furthermore, verapamil may have a direct effect on calcium flux across the sarcolemmal membrane or within intracellular compartments by enhancing calcium homeostasis, which could be protective of reversibly injured myocardial cells, in addition to its vasodilatory effect and its inhibitory effect on platelet aggregation during AMI (Ikeda 1981; Taniyama 1997). Verapamil has been used as an antiarrhythmic agent because of its action on AV nodes (Akhtar 1989; Singh 1987) and AV block. Side effects are more common with verapamil than with adenosine (Vijayalakshmi 2006).

How the intervention might work

Reperfusion‐related injury after PPCI has been associated with high activated neutrophil count, and with activation and aggregation of platelets, and can result in damage to the endothelium, plugging of capillaries, production of oxygen free radicals and release of potent vasoconstrictors such as thromboxane‐A, endothelin‐1 and other inflammatory mediators. These can subsequently contribute to sustained spasm of the coronary microcirculation (Niccoli 2009a). Also, calcium overload can occur in myocytes. To treat patients and prevent the no‐reflow phenomenon, adenosine and verapamil have been used as pharmacological strategies (Niccoli 2010), as they are effective with different mechanisms of no‐reflow by inducing vasodilation, antithrombotic effects and anti‐inflammatory effects, and by decreasing free radical production and calcium overload in myocytes. Adenosine is safe and feasible in people with AMI as an adjunct to PPCI, and it can decrease the incidence of no‐reflow (Marzilli 2000). Similarly, verapamil can reverse no‐reflow after PPCI (Werner 2002).

According to the European Society of Cardiology guidelines for AMI (Van de Werf 2008), adenosine and verapamil have been endorsed as treatments for no‐reflow with a class IIb recommendation at evidence levels of B and C, respectively.

Why it is important to do this review

First, AMI with no‐reflow after PPCI is difficult to treat and is strongly associated with poor clinical in‐hospital and long‐term outcomes (Niccoli 2009a; Niccoli 2010). Second, randomised controlled trials of adenosine and verapamil in people with AMI who are undergoing PPCI have been conducted for many years (Fokkema 2009; Grygier 2011; Marzilli 2000; Micari 2005; Petronio 2005; Taniyama 1997). However, their efficacy for this patient group has not been systematically reviewed to date. This review was undertaken to summarise available evidence and to provide important clinical guidance for management of no‐reflow with AMI after PPCI.

Objectives

To study the impact of adenosine and verapamil on no‐reflow during PPCI in people with AMI.

Methods

Criteria for considering studies for this review

Types of studies

We selected randomised controlled trials only.

Types of participants

We included in our review people diagnosed with AMI and undergoing PPCI, irrespective of age, sex and ethnic group. We also included individuals with suboptimal reperfusion after PPCI. As this review focuses on the impact of adenosine or verapamil on patients given PPCI, we excluded those with AMI receiving emergency thrombolytic therapy before PPCI. As asthma, renal impairment, malignant disease and haemodynamic instability may influence the outcomes of interest, we also excluded studies that included participants with these conditions.

Haemodynamic instability is defined as systolic blood pressure persistently < 90 mmHg or mean arterial pressure < 60 mmHg, or as the need to use inotropic drugs or an intra‐aortic balloon pump to keep systolic pressure at ≥ 90 mmHg.

Types of interventions

Comparison: adenosine or verapamil versus placebo.

We included trials in which adenosine or verapamil was compared with placebo, regardless of dosage, frequency, duration or route of administration (intravenous or intracoronary).

Types of outcome measures

As case‐specific event rates are not available at the same time point for every study, we defined short term as four weeks to six weeks, and defined long term as six months to eighteen months.

Primary outcomes

• All‐cause mortality at short term and at long term.

• Non‐fatal myocardial infarction at short term and at long term.

Secondary outcomes

• Thrombolysis in myocardial infarction (TIMI) flow grade < 3 after PPCI and at short term or at discharge.

• Myocardial blush grade (MBG) 0 to 1 after PPCI and at short term or at discharge.

• Adverse events of specific interest, including hypotension, bradycardia and AV block (all types).

• Economic outcome: treatment costs (US dollars (USD)).

• Quality of life.

Search methods for identification of studies

Electronic searches

We searched for randomised controlled trials in the following electronic databases from the date of last research included in the previously published version of the review in Febuary 2012 to 25 June 2014 for the update.

• Cochrane Central Register of Controlled Trials (CENTRAL) (2014, Issue 5).

• MEDLINE Ovid (1948 to Week 3 June 2014).

• EMBASE Classic and EMBASE Ovid (1947 to 25 June 2014).

• Information for Scientific Information (ISI) Web of Science (1970 to 25 June 2014).

• ISI BIOSIS Previews (1969 to 25 June 2014).

• China National Knowledge Infrastructure (CNKI) (http://www.cnki.net/) (1980 to 25 June 2014).

We have applied the Cochrane RCT filter to MEDLINE (sensitivity‐maximising) and EMBASE, and an adaptation of it to Web of Science and BIOSIS (Lefebvre 2011).

Please see Appendix 1 for search strategies used in this update, and Appendix 2 for details of the original search conducted in 2012.

We applied no language restrictions, and we had non‐English language papers translated for full assessment.

Searching other resources

We searched the following clinical trial registers on 25 June 2014.

• ClinicalTrials.gov (http://www.clinicaltrials.gov/).

• Current Controlled Trials (http://www.controlled‐trials.com/isrctn/).

• Australian and New Zealand Clinical Trials Registry (ANZCTR) (http://www.anzctr.org.au/).

• World Health Organization International Clinical Trials Registry Platform (WHO ICTRP) (http://apps.who.int/trialsearch/).

We handsearched The American Journal of Cardiology from 2000 to 2014.

Data collection and analysis

Selection of studies

Two review authors (NTS and LL) independently screened titles and abstracts. When potentially relevant abstracts were identified, we retrieved the full articles to assess whether they met our inclusion criteria. Review authors resolved disagreements by discussion, or by consultation with a third review author (SQ), until we reached a consensus. Two review authors (LL and SQ) translated Chinese language papers.

Data extraction and management

Two review authors (NTS and SQ) independently extracted data from the trial reports, using a data extraction form. When we could not resolve disagreements by discussion, we consulted a third review author (LL) from our team, or we contacted study authors for clarification. We entered the following information onto the data extraction form.

• General information: record number, trial identification number, date of publication.

• Characteristics of the included trial: design, allocation concealment, duration.

• Participants: diagnosis, numbers of people in experimental and control groups, age, sex, withdrawals/losses to follow‐up (reason/description).

• Interventions: dose, route and timing of administration.

• Outcomes: as defined above, with length of follow‐up.

For binary data, we extracted the total number of participants (denominator) and the number of events (numerator) in each group.

Assessment of risk of bias in included studies

Two review authors (NTS and SQ) independently assessed risk of bias of the included studies under the guidance of the Cochrane Heart Group, by using the tool described in the Cochrane Handbook for Systematic Reviews of Interventions (Cochrane Handbook). We resolved disagreements by discussion, or we contacted study authors to obtain further information. We present risk of bias assessment tables for the following domains (Cochrane Handbook).

• Sequence generation.

• Allocation concealment.

• Blinding.

• Incomplete outcome data.

• Selective outcome reporting.

• Other bias.

Each domain can be categorised as having low, unclear or high risk of bias.

Measures of treatment effect

We analysed extracted data using Review Manager 5 software. For dichotomous outcomes, we used the risk ratio (RR), as it has been shown that the risk ratio is more intuitive than the odds ratio, and that the odds ratio tends to be interpreted by clinicians as the risk ratio (Boissel 1999; Deeks 2000). This misinterpretation can lead to overestimation of the effect. We provide the 95% confidence interval (CI) and the P value for each point estimate. We did not use continuous data in this review.

Unit of analysis issues

We addressed special issues in the analysis of studies using non‐standard design, such as cross‐over trials, cluster‐randomised trials and studies with multiple treatment groups, if they met our inclusion criteria. Petronio 2005 was a multiple treatment group study, from which we extracted only relevant data.

Dealing with missing data

We tried to contact study authors to obtain missing data, and we received a reply from Desmet 2011. We used an imputation method to determine missing data, but our overall findings were not affected, so we have not used the imputation method result in this review (Analysis 1.2).

1.2. Analysis.

Comparison 1 All‐cause mortality, Outcome 2 All‐cause mortality at long term.

Assessment of heterogeneity

We assessed clinical heterogeneity by considering variability in participants, interventions and outcomes of studies. We assessed statistical heterogeneity using the Chi² test and the I² test; we interpreted I² values between 50% and 90% as detecting substantial heterogeneity (Cochrane Handbook; Higgins 2003). The importance of the value of I² depends on the magnitude and direction of effects and the strength of evidence for heterogeneity (e.g. P value from the Chi² test, confidence interval for I²).

Assessment of reporting biases

We had planned to investigate potential publication bias of included studies by constructing a funnel plot (Egger 1997). However, although we include in this review a total of 11 trials, the maximum number included for any single outcome is nine trials, which is deemed insufficient to support a meaningful funnel plot (Cochrane Handbook).

Data synthesis

We combined trial results using the Mantel‐Haenszel risk ratio, fixed‐effect model, except when statistical heterogeneity (I² > 50%) was observed, in which case we used a random‐effects model.

Subgroup analysis and investigation of heterogeneity

If clinical heterogeneity was suspected, we planned subgroup analyses based on the following.

• Participant characteristics (age, sex).

• For adenosine, early administration (within four hours of onset of myocardial infarction symptoms) versus late administration to AMI participants.

• Route of administration.

Sensitivity analysis

We performed sensitivity analyses to assess the impact of variable risks of bias, eligibility criteria, heterogeneity or publication status of included studies. In addition, we compared the combined results of fixed‐effect and random‐effects models when appropriate.

Results

Description of studies

We have provided the details of each included study in the Characteristics of included studies section.

Results of the search

We identified 2177 studies from the search of electronic databases. We identified 105 records via other resources: 37 studies from Clinical Trials.gov; eight from Current Controlled Trials; 15 from ANZCTR; four from WHO ICTRP; 39 from CNKI; and two from handsearching of The American Journal of Cardiology. After de‐duplication, we screened 1448 abstracts, of which 1399 were excluded. We reviewed the remaining 49 articles in full‐text version. We excluded 38 studies, leaving 11 RCTs for inclusion in the review. We found three ongoing studies (Characteristics of ongoing studies). We excluded the four former studies (Huang 2009; Jing 2010; Micari 2004; Zhen‐Guo 2010) that awaited classification, as we could obtain no further information, although we had sent emails to the corresponding authors.

We present the Quality of Reporting Meta‐analysis (QUOROM) flow diagram in Figure 1.

1.

Study flow diagram: 2014 update search (one additional study included).

Included studies

All included studies were conducted with inpatients. Six were carried out in European countries, including the Netherlands (Fokkema 2009; Stoel 2008), Poland (Grygier 2011), Belgium (Desmet 2011) and Italy (Marzilli 2000; Petronio 2005). The remaining five were completed in China (Ji 2007; Wang 2008; Zhang 2012), Japan (Taniyama 1997) and the USA (Micari 2005).

The 11 included RCTs covered 1027 participants, of whom 76% were men. All participants were diagnosed with acute myocardial infarction (AMI) and required primary percutaneous coronary intervention (PPCI). Studies included from 28 to 448 participants.

Of the 11 included trials, 10 compared adenosine versus placebo, and one (Taniyama 1997) compared verapamil versus placebo, as an adjunct to PPCI. The route of administration in all included studies was intracoronary, except for Micari 2005, which administered treatment intravenously. The dose of adenosine ranged from 300 μg to 60 mg, given by infusion or by bolus. Normal saline was used as the placebo in all included trials.

All studies reported outcomes within one month, two trials (Petronio 2005; Zhang 2012) reported six‐month follow‐up and two trials (Desmet 2011; Stoel 2008) reported 12‐month follow‐up.

Excluded studies

Excluded studies are reported in the Characteristics of excluded studies section, along with reasons for exclusion.

Risk of bias in included studies

We have summarised risk of bias results in Figure 2 and Figure 3.

2.

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

3.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Allocation

Although all included studies were described as randomised trials, investigators provided no clear description of randomisation and concealment of allocation, except for Desmet 2011, which reported adequate random sequence generation and adequate concealment of allocation.

Blinding

Details of random sequence generation, concealment and blinding are the most frequent poorly reported parameters in the included trials. In this review, two RCTs stated that they were double‐blind (Desmet 2011; Ji 2007), but only Desmet 2011 provided adequate sequence generation, allocation concealment and blinding. Because of perceived difficulties in blinding, we investigated evidence suggesting that those collecting and analysing outcome data were blinded to group allocation. Five studies reported blinding of this nature (Fokkema 2009; Grygier 2011; Marzilli 2000; Micari 2005; Petronio 2005).

Incomplete outcome data

Most studies did not report missing data and drop‐outs. Desmet 2011 and Fokkema 2009 reported missing data in some outcomes, and Stoel 2008 reported losses to follow‐up. We performed no intention‐to‐treat analyses to account for missing data.

Selective reporting

We have reported in the Results section all outcomes predefined in the Methods section of published trial reports.

Other potential sources of bias

We found no other obvious potential sources of bias.

Effects of interventions

All‐cause mortality

All‐cause mortality at short term (adenosine)

Seven included studies reported mortality at one‐month follow‐up for 830 participants (Desmet 2011; Fokkema 2009; Grygier 2011; Marzilli 2000; Micari 2005; Petronio 2005; Zhang 2012). None demonstrated a clinically or statistically significant difference between adenosine and control groups, with a fixed‐effect model risk ratio (RR) of 0.61 (95% confidence interval (CI) 0.25 to 1.48, P value = 0.27 and I² = 0%) (Analysis 1.1). We found no evidence of a reduction in all‐cause mortality with adenosine at one‐month follow‐up.

1.1. Analysis.

Comparison 1 All‐cause mortality, Outcome 1 All‐cause mortality at short term.

All‐cause mortality at long term (adenosine)

Three studies reported mortality at one‐year follow‐up for 217 participants (Desmet 2011; Stoel 2008; Zhang 2012). We found no evidence of a reduction in all‐cause mortality with adenosine at long‐term follow‐up (fixed‐effect model, RR 0.78, 95% CI 0.22 to 2.74, P value = 0.70, I² = 0%) (Analysis 1.2).

Non‐fatal myocardial infarction at short term (adenosine)

Five studies assessed non‐fatal myocardial infarction at one‐month follow‐up for 658 participants (Fokkema 2009; Grygier 2011; Marzilli 2000; Micari 2005; Zhang 2012). The fixed‐effect model RR was 1.32 (95% CI 0.33 to 5.29, P value = 0.69, I² = 0%) (Analysis 2.1). No evidence currently supports the use of adenosine in reducing non‐fatal myocardial infarction at short‐term follow‐up.

2.1. Analysis.

Comparison 2 Non‐fatal myocardial infarction at short term, Outcome 1 Non‐fatal myocardial infarction.

Thrombolysis in myocardial infarction (TIMI) flow grade < 3 after PPCI (adenosine)

Nine trials with 907 participants reported data on TIMI flow grade < 3 measured after PPCI (Desmet 2011; Fokkema 2009; Grygier 2011; Ji 2007; Marzilli 2000; Micari 2005; Petronio 2005; Wang 2008; Zhang 2012). The fixed‐effect model RR was 0.62 (95% CI 0.42 to 0.91, P value = 0.01, I² = 37%) (Analysis 3.1), indicating that adenosine can reduce the incidence of TIMI flow grade < 3 after PPCI.

3.1. Analysis.

Comparison 3 TIMI flow < 3 after PPCI, Outcome 1 TIMI flow < 3 after PPCI.

Myocardial blush grade (MBG) 0 to 1 after PPCI (adenosine)

Three trials with 621 participants reported MBG 0 to 1 after PPCI (Desmet 2011; Fokkema 2009; Grygier 2011). The fixed‐effect model RR was 0.96 (95% CI 0.76 to 1.22, P value = 0.75, I² = 0%) (Analysis 4.1). Currently no evidence shows a reduction in the incidence of MBG 0 to 1 with adenosine.

4.1. Analysis.

Comparison 4 MBG 0 to 1 after PPCI, Outcome 1 MBG 0 to 1 after PPCI.

Adverse events

Bradycardia (adenosine)

Five included RCTs with 518 people reported on bradycardia after PPCI (Fokkema 2009; Ji 2007; Marzilli 2000; Wang 2008; Zhang 2012). We observed no evidence of statistical heterogeneity between the two groups. The fixed‐effect model RR was 6.32 (95% CI 2.98 to 13.41, P value < 0.00001, I² = 0%) (Analysis 5.1), indicating a significantly greater likelihood of bradycardia in the adenosine group compared with the control group. Bradycardia was defined as heart rate less than 60 beats per minute.

5.1. Analysis.

Comparison 5 Adverse events, Outcome 1 Bradycardia.

Hypotension (adenosine)

Three RCTs with 451 people reported on hypotension after PPCI (Fokkema 2009; Grygier 2011; Marzilli 2000). The fixed‐effect model RR was 11.43 (95% CI 2.75 to 47.57, P value = 0.0008) (Analysis 5.2). Evidence showed increased hypotension after PPCI with adenosine, compared with controls. Hypotension was defined as systolic blood pressure ≤ 90 mmHg.

5.2. Analysis.

Comparison 5 Adverse events, Outcome 2 Hypotension.

Atrioventricular (AV) block (adenosine)

Atrioventricular block after PPCI was reported in six included studies with 639 participants (Desmet 2011; Fokkema 2009; Marzilli 2000; Micari 2005; Petronio 2005; Zhang 2012). As statistical heterogeneity between trials was significant (I² = 50%), we used the random‐effects model (RR 6.78, 95% CI 2.15 to 21.38, P value = 0.001) (Analysis 5.3). Evidence favoured adenosine over control in reducing the occurrence of AV block after PPCI. A sensitivity analysis comparing fixed‐effect and random‐effects models to control for heterogeneity did not substantively alter the results.

5.3. Analysis.

Comparison 5 Adverse events, Outcome 3 AV block.

Economic outcome

No data were available across all randomised controlled trials.

Quality of life

No data were available across all randomised controlled trials.

Discussion

Summary of main results

We have included 11 randomised controlled trials (RCTs) involving 1027 participants with acute myocardial infarction (AMI) for primary percutaneous coronary intervention (PPCI), in which the primary intention was to study the impact of adenosine and verapamil on no‐reflow during PPCI in people with AMI.

Only one of the 11 trials used verapamil (Taniyama 1997) and could not be included in the meta‐analyses. In the Taniyama 1997 study, after verapamil had been administered, no malignant ventricular arrhythmia or haemodynamic abnormalities were observed, and the number of participants with TIMI flow grade < 3 decreased to three from six. However, this study did not report outcomes such as all‐cause mortality, non‐fatal myocardial infarction or myocardial blush grade (MBG) 0 to 1.

We included 10 RCTs involving 987 participants for the study of adenosine. We found no evidence indicating that adenosine reduced short‐term (risk ratio (RR) 0.61, 95% confidence interval (CI) 0.25 to 1.48, P value = 0.27 ) or long‐term (RR 0.78, 95% CI 0.22 to 2.74, P value = 0.70) all‐cause mortality, or non‐fatal myocardial infarction (RR 1.32, 95% CI 0.33 to 5.29, P value = 0.69), to a greater extent than control. Evidence suggests that adenosine can reduce the incidence of TIMI flow grade < 3 after PPCI (RR 0.62, 95% CI 0.42 to 0.91, P value = 0.01), but MBG 0 to 1 is not statistically significant (RR 0.96, 95% CI 0.76 to 1.22, P value = 0.75). Adverse events reported with adenosine include bradycardia (five RCTs), hypotension (four RCTs) and all types of atrioventricular (AV) block (six RCTs). The incidence of adverse events (bradycardia: RR 6.32, 95% CI 2.98 to 13.41, P value < 0.00001; hypotension: RR 11.43, 95% CI 2.75 to 47.57, P value = 0.0008; AV block: RR 6.78, 95% CI 2.15 to 21.38, P value = 0.001)) after the procedure was significantly increased among groups treated with adenosine compared with placebo groups.

Results of the previous meta‐analysis show that adenosine could not significantly reduce the incidence of TIMI flow < 3 after PPCI (RR 0.72, 95% CI 0.49 to 1.07), but a trend toward reduction of TIMI was noted. Unlike the previous analysis, the present update analysis included a study from Zhang et al. Although this study included a small population, it had a large effect on the results of TIMI flow grade < 3 after PPCI. In contrast with the previous meta‐analysis, the present updated meta‐analysis suggests that adenosine is associated with a lower incidence of TIMI flow grade < 3 after PPCI. The present updated meta‐analysis includes one new, small study, which had unclear risk of bias across multiple domains and reported a large effect on the outcome of TIMI flow grade < 3 after PPCI. In fact, other than the study from Fokkema, all included studies were small studies. Risk of small study bias is high, and this has likely influenced the results (Nuesch 2010).

The present updated meta‐analysis had some limitations. First, most included studies had a relatively short period of follow‐up and included a limited number of participants; few studies reported long‐term follow‐up, and event rates of mortality and angiographic no‐reflow were low. The present meta‐analysis therefore had poor statistical power to support clear conclusions about the impact of adenosine in people with AMI undergoing PPCI. Second, compared with the outcome of TIMI flow grade < 3, the summative size of the effect of adenosine on short‐term all‐cause mortality (39% risk reduction) is clinically important, especially in the context of large numbers of individuals with AMI undergoing PPCI. Third, the included new study had unclear risk of bias across multiple domains and high risk of small study bias. Because of these potential clinical benefits, the conclusion should be interpreted cautiously. Large, high‐quality RCTs in this field are needed to explore the potential risks and benefits of these interventions.

Overall completeness and applicability of evidence

In designing this review, we hoped to determine the impact of adenosine or verapamil on AMI by searching for studies that met the following criteria.

• Participants with AMI undergoing PPCI.

• Adenosine or verapamil given as the primary intervention adjunctive to PPCI.

• Key outcomes, including mortality, non‐fatal MI, TIMI flow grade and so forth, reported.

We believe that these parameters are important in helping us answer the clinical question: What is the impact of adenosine or verapamil as an adjunct to PPCI in people with AMI? Extracted and analysed data show that adenosine did not produce a significant reduction in primary outcomes (all‐cause mortality at short term and at long term, and non‐fatal MI at short term). However, we found no data on risks of non‐fatal MI at long term, so we could not investigate this outcome on the basis of available evidence.

According to the guidelines of the European Society of Cardiology, adenosine is recommended as a drug of choice for the treatment of patients with no‐reflow phenomenon, with a IIb class recommendation at evidence level B (Van de Werf 2008). Although no evidence in our previous review suggests that adenosine can reduce the incidence of TIMI flow grade < 3 after PPCI, in this updated review we show that adenosine reduces the incidence of angiographic no‐reflow, and the incidence of MBG 0 to 1 or mortality at short‐term or at long‐term follow‐up remains unchanged. In terms of the change in TIMI flow grade < 3 after PPCI, we found that before this new study (Zhang 2012) was considered, adenosine had already shown a trend towards improving coronary blood flow, but without statistical significance; after addition of the new study, which demonstrated that adenosine could significantly reduce the incidence of TIMI flow grade < 3 after PPCI, we noted the beneficial effects of adenosine in reducing the incidence of TIMI < 3 after PPCI. The change may be a result of the different means of administration and dosage of adenosine used in this new study, in which adenosine was given at 50 or 70 μg/kg/min for 3 hours intravenously. In the light of heterogeneity observed in each study and the small number of included participants, we believe that additional high‐quality RCTs are needed to confirm the efficacy of adenosine for no‐reflow in patients with AMI undergoing PPCI .

Our analyses clearly indicate that the incidence of adverse effects in the adenosine groups (bradycardia, hypotension, AV block) is greater than in the control groups. However, these adverse effects were shown to resolve spontaneously within a few minutes after the procedure.

Participant‐oriented outcomes, such as quality of life and costs of treatment, were seldom reported. Therefore, quality of life and economic outcomes were not included in the analysis. We suggest that future research should focus on participant outcomes.

Quality of the evidence

Although included studies did not describe methods of randomisation and allocation concealment, most studies did report details indicating whether outcomes were assessed by blinded researchers or whether the researchers were independent to the study. So, the overall risk of bias of trials included in this systematic review was judged to be moderate (please refer to Figure 2 and Figure 3).

Potential biases in the review process

We have strictly followed our published protocol in implementing the processes of searching for trials, extracting data and analysing data. However, the possibility of unidentified and/or unpublished papers is a potential source of bias for any systematic review. Another potential risk of bias may be performance bias due to imbalance of co‐interventions across intervention and control arms.

Agreements and disagreements with other studies or reviews

We know of no other systematic reviews of adenosine and verapamil given to individuals with AMI undergoing PPCI.

Authors' conclusions

Implications for practice.

Evidence favours adenosine as an adjunct to PPCI in reducing the incidence of angiographic no‐reflow (TIMI flow grade < 3), but no evidence shows reduced mortality. Although the incidence of adverse effects of adenosine (hypotension, bradycardia, AV block) is statistically significant, the effects are short‐lived.

In accordance with the current review, and in keeping with international guidelines (Steg 2012; Van de Werf 2008), we believe that adenosine as adjunctive therapy for patients with AMI undergoing PPCI should not be excluded from clinical practice.

Implications for research.

Although primary outcomes (all‐cause mortality and non‐fatal MI) showed no significant differences between adenosine and control groups, this evidence was based on relatively few included RCTs with few participants that were not powered to deliver a definitive judgement on adenosine in people with AMI. We hope that future RCTs can focus on both short‐term and long‐term clinical outcomes (all‐cause mortality and non‐fatal MI). Our finding is subject to small study bias, and conclusions should be interpreted cautiously.

It is worth noting that myocardial blush grade and thrombolysis in myocardial infarction flow grade are important outcomes for assessment of angiographic no‐reflow after PPCI. Although adenosine has been shown to be superior to placebo as treatment for angiographic no‐reflow after PPCI, the myocardial blush grade is currently limited. Larger trials of high quality are needed. Because no available data can be used, quality of life and health economic outcomes are not included in the analysis. Future trials should incorporate quality of life and health economic outcomes.

This review identified only one trial of verapamil for inclusion. Additional RCTs testing verapamil as an adjunct to PPCI in people with AMI are necessary, to explore its potential as adjunctive treatment.

What's new

Date	Event	Description
8 July 2014	New search has been performed	This update includes 1 new study.
25 June 2014	New citation required and conclusions have changed	Results of TIMI flow grade < 3 after PPCI have changed after a new study was added (Zhang 2012). The rest of our conclusions remain the same.

Appendices

Appendix 1. Search strategies 2014

CENTRAL

#1 MeSH descriptor Vasodilator Agents, this term only
 #2 (vasodilator*)
 #3 (vasorelaxant*)
 #4 ("vasoactive antagon*")
 #5 MeSH descriptor Adenosine, this term only
 #6 MeSH descriptor Nucleosides, this term only
 #7 (#6), from 1966 to 1971
 #8 (adenosine)
 #9 (adenocard)
 #10 MeSH descriptor Verapamil, this term only
 #11 (verapamil)
 #12 finoptin
 #13 (iproveratril)
 #14 (i?optin*)
 #15 (lekoptin)
 #16 (dexverapamil)
 #17 (calan)
 #18 (falicard)
 #19 (cordilox)
 #20 (adenocor)
 #21 (vertab)
 #22 (securon)
 #23 (univer)
 #24 (verapress)
 #25 (adeno‐jec)
 #26 (adenoscan)
 #27 (apo‐verap)
 #28 (berkatens)
 #29 (chronovera)
 #30 (covera)
 #31 "ethimil mr"
 #32 (finoptin)
 #33 geangin
 #34 (hypaneze)
 #35 "ipoveratril hydrochloride"
 #36 (iproveratril)
 #37 (novo‐veramil)
 #38 (nu‐verap)
 #39 (verelan)
 #40 (#1 OR #2 OR #3 OR #4 OR #5 OR #7 OR #8 OR #9 OR #10 OR #11 OR #12 OR #13 OR #14 OR #15 OR #16 OR #17 OR #18 OR #19 OR #20 OR #21 OR #22 OR #23 OR #24 OR #25 OR #26 OR #27 OR #28 OR #29 OR #30 OR #31 OR #32 OR #33 OR #34 OR #35 OR #36 OR #37 OR #38 OR #39)
 #41 MeSH descriptor Angioplasty, Balloon, Coronary, this term only
 #42 (coronary near/5 angioplast*)
 #43 (pci)
 #44 "percutaneous coronary intervention*"
 #45 (ptca)
 #46 (balloon near/3 angioplast*)
 #47 (coronary near/5 balloon dilation*)
 #48 (coronary near/5 stent*)
 #49 (#41 OR #42 OR #43 OR #44 OR #45 OR #46 OR #47 OR #48)
 #50 (#40 AND #49)

MEDLINE Ovid

1 Vasodilator Agents/
 2 vasodilator*.tw.
 3 vasorelaxant*.tw.
 4 (vasoactive adj antagon*).tw.
 5 Adenosine/
 6 NUCLEOSIDES/
 7 limit 6 to yr="1966 ‐ 1971"
 8 adenosine.tw.
 9 adenocard.tw.
 10 Verapamil/
 11 verapamil.tw.
 12 finoptin.tw.
 13 iproveratril.tw.
 14 i?optin*.tw.
 15 lekoptin.tw.
 16 dexverapamil.tw.
 17 calan.tw.
 18 falicard.tw.
 19 cordilox.tw.
 20 adenocor.tw.
 21 vertab.tw.
 22 securon.tw.
 23 univer.tw.
 24 verapress.tw.
 25 adeno‐jec.tw.
 26 adenoscan.tw.
 27 apo‐verap.tw.
 28 berkatens.tw.
 29 chronovera.tw.
 30 ethimil mr.tw.
 31 finoptin.tw.
 32 geangin.tw.
 33 hypaneze.tw.
 34 ipoveratril hydrochloride.tw.
 35 iproveratril.tw.
 36 novo‐veramil.tw.
 37 nu‐verap.tw.
 38 Angioplasty, Balloon, Coronary/
 39 (coronary adj5 angioplast*).tw.
 40 pci.tw.
 41 percutaneous coronary intervention*.tw.
 42 ptca.tw.
 43 (balloon adj3 angioplast*).tw.
 44 (coronary adj5 balloon dilation*).tw.
 45 (coronary adj5 stent*).tw.
 46 or/38‐45
 47 randomized controlled trial.pt.
 48 controlled clinical trial.pt.
 49 randomized.ab.
 50 placebo.ab.
 51 drug therapy.fs.
 52 randomly.ab.
 53 trial.ab.
 54 groups.ab.
 55 47 or 48 or 49 or 50 or 51 or 52 or 53 or 54
 56 exp animals/ not humans.sh.
 57 55 not 56
 58 covera.tw.
 59 verelan.tw.
 60 1 or 2 or 3 or 4 or 5 or 7 or 8 or 9 or 10 or 11 or 12 or 13 or 14 or 15 or 16 or 17 or 18 or 19 or 20 or 21 or 22 or 23 or 24 or 25 or 26 or 27 or 28 or 29 or 30 or 31 or 32 or 33 or 34 or 35 or 36 or 37 or 58 or 59
 61 46 and 57 and 60

EMBASE

1 Vasodilator Agents/
 2 vasodilator*.tw.
 3 vasorelaxant*.tw.
 4 (vasoactive adj antagon*).tw.
 5 Adenosine/
 6 NUCLEOSIDES/
 7 limit 6 to yr="1966 ‐ 1971"
 8 adenosine.tw.
 9 adenocard.tw.
 10 Verapamil/
 11 verapamil.tw.
 12 finoptin.tw.
 13 iproveratril.tw.
 14 i?optin*.tw.
 15 lekoptin.tw.
 16 dexverapamil.tw.
 17 calan.tw.
 18 falicard.tw.
 19 cordilox.tw.
 20 adenocor.tw.
 21 vertab.tw.
 22 securon.tw.
 23 univer.tw.
 24 verapress.tw.
 25 adeno‐jec.tw.
 26 adenoscan.tw.
 27 apo‐verap.tw.
 28 berkatens.tw.
 29 chronovera.tw.
 30 ethimil mr.tw.
 31 finoptin.tw.
 32 geangin.tw.
 33 hypaneze.tw.
 34 ipoveratril hydrochloride.tw.
 35 iproveratril.tw.
 36 novo‐veramil.tw.
 37 nu‐verap.tw.
 38 Angioplasty, Balloon, Coronary/
 39 (coronary adj5 angioplast*).tw.
 40 pci.tw.
 41 percutaneous coronary intervention*.tw.
 42 ptca.tw.
 43 (balloon adj3 angioplast*).tw.
 44 (coronary adj5 balloon dilation*).tw
 45 (coronary adj5 stent*).tw.
 46 or/38‐45
 47 covera.tw.
 48 verelan.tw.
 49 1 or 2 or 3 or 4 or 5 or 7 or 8 or 9 or 10 or 11 or 12 or 13 or 14 or 15 or 16 or 17 or 18 or 19 or 20 or 21 or 22 or 23 or 24 or 25 or 26 or 27 or 28 or 29 or 30 or 31 or 32 or 33 or 34 or 35 or 36 or 37 or 47 or 48
 50 random$.tw.
 51 factorial$.tw.
 52 crossover$.tw.
 53 cross over$.tw.
 54 cross‐over$.tw.
 55 placebo$.tw.
 56 (doubl$ adj blind$).tw.
 57 (singl$ adj blind$).tw.
 58 assign$.tw.
 59 allocat$.tw.
 60 volunteer$.tw.
 61 crossover procedure/
 62 double blind procedure/
 63 randomized controlled trial/
 64 single blind procedure/
 65 50 or 51 or 52 or 53 or 54 or 55 or 56 or 57 or 58 or 59 or 60 or 61 or 62 or 63 or 64
 66 (animal/ or nonhuman/) not human/
 67 65 not 66
 68 46 and 49 and 67

Web of Science and BIOSIS

#21 #20 AND #19
 #20 Topic=(((random* or blind* or allocat* or assign* or trial* or placebo* or crossover* or cross‐over*))) ;
 #19 #18 AND #12 ;
 #18 #17 OR #16 OR #15 OR #14 OR #13
 #17 Topic=((coronary near/5 stent*))
 #16 Topic=(((coronary )near/5 ("balloon dilation*")))
 #15 Topic=(("percutaneous coronary intervention*") or (ptca) or (balloon near/3 angioplast*)) ;
 #14 Topic=(pci)
 #13 Topic=((coronary near/5 angioplast*))
 #12 #11 OR #10 OR #9 OR #8 OR #7 OR #6 OR #5 OR #4 OR #3 OR #2 OR #1
 #11 Topic=((iproveratril) or (novo‐veramil) or (nu‐verap) or (verelan))
 #10 Topic=((finoptin) or (geangin) or (hypaneze) or ("ipoveratril hydrochloride"))
 #9 Topic=((berkatens) or (chronovera) or (covera) or (ethimil))
 #8 Topic=((verapress) or (adeno‐jec) or (adenoscan) or (apo‐verap))
 #7 Topic=((adenocor) or (vertab) or (securon) or (univer))
 #6 Topic=((calan) or (falicard) or (cordilox)) ;
 #5 Topic=((i?optin*) or (lekoptin) or (dexverapamil))
 #4 Topic=((verapamil) or ( finoptin) or ( iproveratril))
 #3 Topic=(("vasoactive antagon*") or (adenosine) or (adenocard))
 #2 Topic=((vasorelaxant*))
 #1 Topic=((vasodilator*))

CNKI

1. 腺苷 and 急性心肌梗死

2. 维拉帕米 and 急性心肌梗死

Clinical Trial Registers

1. adenosine and myocardial infarction
 2. verapamil and myocardial infarction

Appendix 2. Search strategies 2012

CENTRAL

#1 MeSH descriptor Vasodilator Agents, this term only
 #2 (vasodilator*)
 #3 (vasorelaxant*)
 #4 ("vasoactive antagon*")
 #5 MeSH descriptor Adenosine, this term only
 #6 MeSH descriptor Nucleosides, this term only
 #7 (#6), from 1966 to 1971
 #8 (adenosine)
 #9 (adenocard)
 #10 MeSH descriptor Verapamil, this term only
 #11 (verapamil)
 #12 finoptin
 #13 (iproveratril)
 #14 (i?optin*)
 #15 (lekoptin)
 #16 (dexverapamil)
 #17 (calan)
 #18 (falicard)
 #19 (cordilox)
 #20 (adenocor)
 #21 (vertab)
 #22 (securon)
 #23 (univer)
 #24 (verapress)
 #25 (adeno‐jec)
 #26 (adenoscan)
 #27 (apo‐verap)
 #28 (berkatens)
 #29 (chronovera)
 #30 (covera)
 #31 "ethimil mr"
 #32 (finoptin)
 #33 geangin
 #34 (hypaneze)
 #35 "ipoveratril hydrochloride"
 #36 (iproveratril)
 #37 (novo‐veramil)
 #38 (nu‐verap)
 #39 (verelan)
 #40 (#1 OR #2 OR #3 OR #4 OR #5 OR #7 OR #8 OR #9 OR #10 OR #11 OR #12 OR #13 OR #14 OR #15 OR #16 OR #17 OR #18 OR #19 OR #20 OR #21 OR #22 OR #23 OR #24 OR #25 OR #26 OR #27 OR #28 OR #29 OR #30 OR #31 OR #32 OR #33 OR #34 OR #35 OR #36 OR #37 OR #38 OR #39)
 #41 MeSH descriptor Angioplasty, Balloon, Coronary, this term only
 #42 (coronary near/5 angioplast*)
 #43 (pci)
 #44 "percutaneous coronary intervention*"
 #45 (ptca)
 #46 (balloon near/3 angioplast*)
 #47 (coronary near/5 balloon dilation*)
 #48 (coronary near/5 stent*)
 #49 (#41 OR #42 OR #43 OR #44 OR #45 OR #46 OR #47 OR #48)
 #50 (#40 AND #49)

MEDLINE Ovid

1 Vasodilator Agents/
 2 vasodilator*.tw.
 3 vasorelaxant*.tw.
 4 (vasoactive adj antagon*).tw.
 5 Adenosine/
 6 NUCLEOSIDES/
 7 limit 6 to yr="1966 ‐ 1971"
 8 adenosine.tw.
 9 adenocard.tw.
 10 Verapamil/
 11 verapamil.tw.
 12 finoptin.tw.
 13 iproveratril.tw.
 14 i?optin*.tw.
 15 lekoptin.tw.
 16 dexverapamil.tw.
 17 calan.tw.
 18 falicard.tw.
 19 cordilox.tw.
 20 adenocor.tw.
 21 vertab.tw.
 22 securon.tw.
 23 univer.tw.
 24 verapress.tw.
 25 adeno‐jec.tw.
 26 adenoscan.tw.
 27 apo‐verap.tw.
 28 berkatens.tw.
 29 chronovera.tw.
 30 ethimil mr.tw.
 31 finoptin.tw.
 32 geangin.tw.
 33 hypaneze.tw.
 34 ipoveratril hydrochloride.tw.
 35 iproveratril.tw.
 36 novo‐veramil.tw.
 37 nu‐verap.tw.
 38 Angioplasty, Balloon, Coronary/
 39 (coronary adj5 angioplast*).tw.
 40 pci.tw.
 41 percutaneous coronary intervention*.tw.
 42 ptca.tw.
 43 (balloon adj3 angioplast*).tw.
 44 (coronary adj5 balloon dilation*).tw.
 45 (coronary adj5 stent*).tw.
 46 or/38‐45
 47 randomized controlled trial.pt.
 48 controlled clinical trial.pt.
 49 randomized.ab.
 50 placebo.ab.
 51 drug therapy.fs.
 52 randomly.ab.
 53 trial.ab.
 54 groups.ab.
 55 47 or 48 or 49 or 50 or 51 or 52 or 53 or 54
 56 exp animals/ not humans.sh.
 57 55 not 56
 58 covera.tw.
 59 verelan.tw.
 60 1 or 2 or 3 or 4 or 5 or 7 or 8 or 9 or 10 or 11 or 12 or 13 or 14 or 15 or 16 or 17 or 18 or 19 or 20 or 21 or 22 or 23 or 24 or 25 or 26 or 27 or 28 or 29 or 30 or 31 or 32 or 33 or 34 or 35 or 36 or 37 or 58 or 59
 61 46 and 57 and 60

EMBASE

1 Vasodilator Agents/
 2 vasodilator*.tw.
 3 vasorelaxant*.tw.
 4 (vasoactive adj antagon*).tw.
 5 Adenosine/
 6 NUCLEOSIDES/
 7 limit 6 to yr="1966 ‐ 1971"
 8 adenosine.tw.
 9 adenocard.tw.
 10 Verapamil/
 11 verapamil.tw.
 12 finoptin.tw.
 13 iproveratril.tw.
 14 i?optin*.tw.
 15 lekoptin.tw.
 16 dexverapamil.tw.
 17 calan.tw.
 18 falicard.tw.
 19 cordilox.tw.
 20 adenocor.tw.
 21 vertab.tw.
 22 securon.tw.
 23 univer.tw.
 24 verapress.tw.
 25 adeno‐jec.tw.
 26 adenoscan.tw.
 27 apo‐verap.tw.
 28 berkatens.tw.
 29 chronovera.tw.
 30 ethimil mr.tw.
 31 finoptin.tw.
 32 geangin.tw.
 33 hypaneze.tw.
 34 ipoveratril hydrochloride.tw.
 35 iproveratril.tw.
 36 novo‐veramil.tw.
 37 nu‐verap.tw.
 38 Angioplasty, Balloon, Coronary/
 39 (coronary adj5 angioplast*).tw.
 40 pci.tw.
 41 percutaneous coronary intervention*.tw.
 42 ptca.tw.
 43 (balloon adj3 angioplast*).tw.
 44 (coronary adj5 balloon dilation*).tw
 45 (coronary adj5 stent*).tw.
 46 or/38‐45
 47 covera.tw.
 48 verelan.tw.
 49 1 or 2 or 3 or 4 or 5 or 7 or 8 or 9 or 10 or 11 or 12 or 13 or 14 or 15 or 16 or 17 or 18 or 19 or 20 or 21 or 22 or 23 or 24 or 25 or 26 or 27 or 28 or 29 or 30 or 31 or 32 or 33 or 34 or 35 or 36 or 37 or 47 or 48
 50 random$.tw.
 51 factorial$.tw.
 52 crossover$.tw.
 53 cross over$.tw.
 54 cross‐over$.tw.
 55 placebo$.tw.
 56 (doubl$ adj blind$).tw.
 57 (singl$ adj blind$).tw.
 58 assign$.tw.
 59 allocat$.tw.
 60 volunteer$.tw.
 61 crossover procedure/
 62 double blind procedure/
 63 randomized controlled trial/
 64 single blind procedure/
 65 50 or 51 or 52 or 53 or 54 or 55 or 56 or 57 or 58 or 59 or 60 or 61 or 62 or 63 or 64
 66 (animal/ or nonhuman/) not human/
 67 65 not 66
 68 46 and 49 and 67

Web of Science and BIOSIS

#21 #20 AND #19
 #20 Topic=(((random* or blind* or allocat* or assign* or trial* or placebo* or crossover* or cross‐over*))) ;
 #19 #18 AND #12 ;
 #18 #17 OR #16 OR #15 OR #14 OR #13
 #17 Topic=((coronary near/5 stent*))
 #16 Topic=(((coronary )near/5 ("balloon dilation*")))
 #15 Topic=(("percutaneous coronary intervention*") or (ptca) or (balloon near/3 angioplast*)) ;
 #14 Topic=(pci)
 #13 Topic=((coronary near/5 angioplast*))
 #12 #11 OR #10 OR #9 OR #8 OR #7 OR #6 OR #5 OR #4 OR #3 OR #2 OR #1
 #11 Topic=((iproveratril) or (novo‐veramil) or (nu‐verap) or (verelan))
 #10 Topic=((finoptin) or (geangin) or (hypaneze) or ("ipoveratril hydrochloride"))
 #9 Topic=((berkatens) or (chronovera) or (covera) or (ethimil))
 #8 Topic=((verapress) or (adeno‐jec) or (adenoscan) or (apo‐verap))
 #7 Topic=((adenocor) or (vertab) or (securon) or (univer))
 #6 Topic=((calan) or (falicard) or (cordilox)) ;
 #5 Topic=((i?optin*) or (lekoptin) or (dexverapamil))
 #4 Topic=((verapamil) or ( finoptin) or ( iproveratril))
 #3 Topic=(("vasoactive antagon*") or (adenosine) or (adenocard))
 #2 Topic=((vasorelaxant*))
 #1 Topic=((vasodilator*))

CNKI

1. 腺苷 and 急性心肌梗死

2. 维拉帕米 and 急性心肌梗死

Clinical Trial Registers

1. adenosine and myocardial infarction
 2. verapamil and myocardial infarction

Data and analyses

Comparison 1. All‐cause mortality.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 All‐cause mortality at short term	7	830	Risk Ratio (M‐H, Fixed, 95% CI)	0.61 [0.25, 1.48]
2 All‐cause mortality at long term	3	217	Risk Ratio (M‐H, Fixed, 95% CI)	0.78 [0.22, 2.74]

Comparison 2. Non‐fatal myocardial infarction at short term.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Non‐fatal myocardial infarction	5	658	Risk Ratio (M‐H, Fixed, 95% CI)	1.32 [0.33, 5.29]

Comparison 3. TIMI flow < 3 after PPCI.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 TIMI flow < 3 after PPCI	9	907	Risk Ratio (M‐H, Fixed, 95% CI)	0.62 [0.42, 0.91]

Comparison 4. MBG 0 to 1 after PPCI.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 MBG 0 to 1 after PPCI	3	621	Risk Ratio (M‐H, Fixed, 95% CI)	0.96 [0.76, 1.22]

Comparison 5. Adverse events.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Bradycardia	5	518	Risk Ratio (M‐H, Fixed, 95% CI)	6.32 [2.98, 13.41]
2 Hypotension	3	452	Risk Ratio (M‐H, Fixed, 95% CI)	11.43 [2.75, 47.57]
3 AV block	6	639	Risk Ratio (M‐H, Random, 95% CI)	6.78 [2.15, 21.38]

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Desmet 2011.

Methods	Date of publication: 2011 Design: prospective, single‐centre, double‐blind, placebo‐controlled clinical study Allocation: randomised Country: Belgium
Participants	Diagnosis: acute STEMI for PPCI Number: 110 (adenosine 56; control 54) Age, years: adenosine (mean 61.4 ± 12.3); control (mean 60.6 ± 10.8) Sex: 82% men Inclusion: chest pain suggestive of myocardial ischaemia for at least 20 minutes; time from onset of symptoms of 12 hours; ECG showing ST‐segment elevation of 0.1 mV in 2 or more limb leads or 0.2 mV in 2 or more contiguous precordial leads, presumed new left bundle branch block Exclusion: contraindication 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; participation in any investigational drug or device study within the previous 30 days Withdrawals or losses to follow‐up: not reported
Interventions	Adenosine Dose: high‐dose bolus (4 mg in 5 mL of 0.9% NaCl) Route of administration: intracoronary Timing: during PPCI, after coronary angiography Control Dose: high‐dose bolus (5 mL of 0.9% NaCl) Route of administration: intracoronary Timing: during PPCI, after coronary angiography
Outcomes	All‐cause death at short term and at long term TIMI (RR 1.52, 95% Cl 0.63 to 3.62) MBG Side effects: bradycardia, AV block Treatment cost: not reported Quality of life: not reported Follow‐up: 30 days and 1 year (collected from hospital records and telephone interviews)
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"Study drug was constituted by the hospital pharmacy in numbered vials according to a computer‐generated randomization list"
Allocation concealment (selection bias)	Low risk	"This list was kept in a sealed envelope in the hospital pharmacy"
Blinding (performance bias and detection bias) All outcomes	Low risk	Double‐blinded: "Vials containing adenosine and those containing placebo had an identical appearance. All study personnel were blinded to treatment allocation until the study had been completed and all analyses had been performed" "Coronary angiograms obtained before and after primary PCI were reviewed by two experienced observers blinded to treatment allocation and clinical data"
Incomplete outcome data (attrition bias) All outcomes	Low risk	We found no data on MBG and no clear description of it in the full text. However, communication with the study author confirmed that data for MBG were missing because of technical problems involved in determining MBG (e.g. when the image was focused on the coronary lesion, not including the myocardial region of interest needed to determine MBG, or when the coronary artery was not filmed in such a way that MBG could be assessed)
Selective reporting (reporting bias)	Low risk	All outcomes stated in Methods are reported in Results
Other bias	Low risk	Not obvious

Fokkema 2009.

Methods	Date of publication: 2009 Design: not reported Allocation: randomised Country: Netherlands
Participants	Diagnosis: acute STEMI for PPCI Number: 448 (adenosine 226; control 222) Age, years: adenosine (mean 62.4 ± 12.2); control (mean 62.5 ± 13.0) Sex: 75% men Inclusion: chest pain suggestive of myocardial ischaemia for at least 30 minutes; time from onset of symptoms of 12 hours before hospital admission; and ECG showing ST‐segment elevation of 0.1 mV in 2 or more leads Exclusion: cardiogenic shock; presence of a life‐threatening disease with a life expectancy of 6 months; receiving pharmacotherapy for chronic obstructive pulmonary disease; no informed consent Withdrawals or losses to follow‐up: not reported
Interventions	Adenosine Dose: 2 high‐dose boluses (2 × 120 μg in 20 mL 0.9% NaCl) Route of administration: intracoronary Timing: during PPCI, after thrombus aspiration and after stenting Control Dose: 2 × 20 mL 0.9% NaCl Route of administration: intracoronary Timing: during PPCI, after thrombus aspiration and after stenting
Outcomes	All‐cause death at short term Non‐fatal MI (reinfarction) TIMI (RR 0.92, 95% Cl 0.44 to 1.90) MBG Side effects: bradycardia, hypotension, AV block Treatment cost: not reported Quality of life: not reported Follow‐up: 30 days (collected from hospital records, written questionnaires and telephone interviews)
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Randomised, no further description
Allocation concealment (selection bias)	High risk	"a single‐center, prospective randomized open trial with blinded evaluation of end points"
Blinding (performance bias and detection bias) All outcomes	Low risk	"Coronary angiograms were reviewed by 2 observers blinded for treatment allocation and clinical data" "QuBE was measured blinded to clinical data and treatment allocation by 2 observers"
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	No withdrawals or drop‐outs were reported, but for some outcomes not all participant data were addressed, with no reason given. Missing outcome data were balanced across groups
Selective reporting (reporting bias)	Low risk	All outcomes stated in Methods are reported in Results
Other bias	Low risk	Not obvious

Grygier 2011.

Methods	Date of publication: 2011 Design: not reported Allocation: randomised Country: Poland
Participants	Diagnosis: acute STEMI for PPCI Number: 70 (adenosine 35; control 35) Age, years: adenosine (mean 65.3 ± 11), control (mean 64.5 ± 13) Sex: 63% men Inclusion: referred for PPCI within 6 hours of symptoms; culprit lesion was suitable for PCI and had presented with TIMI flow 0 to 2 Exclusion: chronic obstructive pulmonary disease or asthma; those who had received previous thrombolysis; patients presenting with TIMI flow 3 Withdrawals or losses to follow‐up: not reported
Interventions	Adenosine Dose: 2 mg to the left or 1 mg to the right coronary artery in 10 mL 0.9% NaCl × 2 times Route of administration: intracoronary Timing: during PPCI, immediately after crossing the lesion by guide wire and after the first balloon inflation Control Dose: 10 mL 0.9% NaCl × 2 times Route of administration: intracoronary Timing: during PPCI, immediately after crossing the lesion by guide wire and after the first balloon inflation
Outcomes	Death Non‐fatal MI TIMI (RR 0.38, 95% Cl 0.11 to 1.30) MBG Side effects: bradycardia, hypotension, AV block (not useable) Follow‐up: 1 month
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	"the patient was included in the present study and randomized after informed consent had been given"; no further description
Allocation concealment (selection bias)	Unclear risk	Not stated
Blinding (performance bias and detection bias) All outcomes	Low risk	"Angiograms were reanalyzed as a single group by 2 observers, who had not participated in the invasive procedure and were unaware of the treatment received" "MBG was assessed by 2 observers who were unaware of the others' findings and of the clinical data" "physicians in charge of the patients on the coronary care floor were informed of the angiographic results of the primary PCI but were unaware of the study treatment administered during the procedure" "The electrocardiograms were analysed in a blinded fashion by an experienced cardiologist"
Incomplete outcome data (attrition bias) All outcomes	Low risk	No incomplete outcome data
Selective reporting (reporting bias)	Low risk	All outcomes stated in Methods are reported in Results
Other bias	Low risk	Not obvious

Ji 2007.

Methods	Date of publication: 2007 Design: double‐blinded Allocation: randomised Country: China
Participants	Diagnosis: acute STEMI for PPCI Number: 50 (adenosine 23; control 27) Age, years: adenosine (mean 59.35 ± 12.06), control (mean 60.59 ± 11.24) Sex: 82% men Inclusion: chest pain persisting at Ieast 30 minutes but no longer than 12 hours; ECG ST‐segment elevation (> 0.1 mV) of at Ieast 2 contiguous ECG leads; cuIprIt Iesion was suitabIe for PCI; informed consent Exclusion: cardiogenic shock or severe heart failure (Killip cIassification ≥ 3); received thromboIytics in the emergency room; history of bronchospasm or undergoing therapy with theophyline; presented with TIMI 3 flow; heart rate < 50 bpm; bIood pressure < 90/60 mmHg Withdrawals or losses to follow‐up: not reported
Interventions	Adenosine Dose: 300 μg in 10 mL saline Route of administration: intracoronary Timing: during PPCI, after balloon dilatation Control Dose: 10 mL saline Route of administration: intracoronary Timing: during PPCI, after balloon inflation and deflation
Outcomes	TIMI flow (RR 0.94, 95% Cl 0.29 to 3.09) Side effects: bradycardia Follow‐up: 4 weeks
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Randomised, no further description
Allocation concealment (selection bias)	Unclear risk	Not described
Blinding (performance bias and detection bias) All outcomes	Unclear risk	Double‐blinded, but no clear statement
Incomplete outcome data (attrition bias) All outcomes	Low risk	No incomplete outcome data
Selective reporting (reporting bias)	Low risk	All outcomes stated in Methods are reported in Results
Other bias	Low risk	Not obvious

Marzilli 2000.

Methods	Date of publication: 2000 Design: not reported Allocation: randomised Country: Italy
Participants	Diagnosis: AMI for PPCI Number: 54 (adenosine 27; control 27) Age, years: adenosine (mean 58.15 ± 11), control (mean 61.9 ± 9) Sex: 80% men Inclusion: referred for PTCA within 3 hours from the onset of AMI; culprit lesion was suitable for PPCI; presented with TIMI flow from 0 to 2; informed consent Exclusion: TIMI 3 flow and having spontaneous reperfusion; history of bronchospasm and/or undergoing therapy with theophylline derivatives; had received thrombolytics in the emergency room Withdrawals or losses to follow‐up: not reported
Interventions	Adenosine Dose: 4 mg in 2 mL saline Route of administration: intracoronary Timing: during PPCI, after balloon dilatation Control Dose: 2 mL saline Route of administration: intracoronary Timing: during PPCI, after balloon dilatation
Outcomes	Death Non‐fatal MI (reinfarction) TIMI (RR 0.06, 95% Cl 0.00 to 0.97) Side effects: bradycardia, hypotension, AV block Duration: during hospitalisation, no further follow‐up
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	"the patient was included in the study and randomised after informed consent had been obtained"; no further description
Allocation concealment (selection bias)	Unclear risk	Not stated
Blinding (performance bias and detection bias) All outcomes	Low risk	"Angiograms were reread as a single group by observers who had not participated in the invasive procedure and were blinded to the treatment received" "Physicians in charge of the patients in the intensive coronary care unit were informed of the angiographic results of the PTCA but were blinded to the intracoronary treatment administered during the procedure" "Images were recorded on S‐VHS tapes and analysed offline by 2 experienced observers blinded to the angiographic data"
Incomplete outcome data (attrition bias) All outcomes	Low risk	No incomplete outcome data
Selective reporting (reporting bias)	Low risk	All measured outcomes were reported in Results
Other bias	Low risk	Not obvious

Micari 2005.

Methods	Date of publication: 2005 Design: not reported Allocation: randomised Country: USA
Participants	Diagnosis: acute STEMI for PPCI Number: 30 (adenosine 14; control 16) Age, years: adenosine (mean 57), control (mean 57) Sex: 67% men Inclusion: symptoms consistent with myocardial ischaemia for 30 minutes and 2‐mm ST‐segment elevation in 2 contiguous electrocardiographic leads; STEMI presenting within 6 hours from the onset of AMI Exclusion: history of AMI; wall motion abnormalities in 1 vascular territory; cardiomyopathy Withdrawals or losses to follow‐up: not reported
Interventions	Adenosine Dose: 50 to 70 μg/kg/min for 3 hours Route of administration: intravenous Timing: initiated < 15 minutes before the procedure Control Dose: normal saline for 3 hours Route of administration: intravenous Timing: initiated < 15 minutes before the procedure
Outcomes	Death Non‐fatal MI (no recurrent ischaemic events) TIMI (RR 1.14, 95% Cl 0.18 to 7.08) Side effects: AV block Follow‐up: 4 weeks
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	"Patients with ST‐elevation AMI referred for percutaneous coronary stenting were randomized to receive intravenous adenosine"; no further description
Allocation concealment (selection bias)	Unclear risk	Not stated
Blinding (performance bias and detection bias) All outcomes	Low risk	"The thrombolysis In myocardial infarction flow grade after PCI was assessed by a reader blinded to the clinical and echocardiographic data" "Analysis of the myocardial contrast echocardiographic data was performed blinded to the clinical and angiographic data" "The analysis of wall motion index was performed blinded to all other data"
Incomplete outcome data (attrition bias) All outcomes	Low risk	No missing data were reported
Selective reporting (reporting bias)	Low risk	All outcomes stated in Methods are reported in Results
Other bias	Low risk	Not obvious

Petronio 2005.

Methods	Date of publication: 2005 Design: not reported Allocation: randomised Country: Italy
Participants	Diagnosis: AMI for PPCI Number: 90 (adenosine 30; abciximab 30; control 30) Age, years: adenosine (mean 56 ± 14), abciximab (mean 60 ± 11), control (mean 61 ± 13) Sex: 83% men Inclusion: presentation > 6 hours from symptom onset; chest pain lasting 30 minutes and resistant to nitrates; 0.2‐mV ST‐segment elevation in 2 contiguous leads on a 12‐lead electrocardiogram; TIMI flow 0 to 1 in the infarct‐related artery at the diagnostic angiogram; absence of contraindications to abciximab and adenosine Exclusion: significant left main coronary disease; cardiogenic shock; AMI due to bypass graft occlusion; thrombolytic therapy before angioplasty Withdrawals or losses to follow‐up: not reported
Interventions	Adenosine Dose: 4 mg Route of administration: intracoronary Timing: during PPCI, before the first balloon dilation Control Dose: normal saline Route of administration: intracoronary Timing: during PPCI, before the first balloon dilation
Outcomes	Death at short term, but no report for long term TIMI (RR 0.50, 95% Cl 0.10 to 2.53) Side effects: AV block Follow‐up: 1 month and 6 months
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	"Randomization occurred in a sequential alternating fashion based on the order of admission at our laboratory"
Allocation concealment (selection bias)	Unclear risk	Not stated.
Blinding (performance bias and detection bias) All outcomes	Low risk	"The angiograms were reread as a single group, in chronological order, by experienced observers blinded to patient treatment" "Interpretation of the echocardiographic was blinded to patient treatment" "The ECGs were analysed in a blinded fashion by an experienced cardiologist"
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	No incomplete outcome data
Selective reporting (reporting bias)	Unclear risk	All outcomes stated in Methods are reported in Results
Other bias	Unclear risk	Not obvious

Stoel 2008.

Methods	Date of publication: 2008 Design: not reported Allocation: randomised Country: Netherlands
Participants	Diagnosis: acute STEMI for PPCI Number: 49 (adenosine 27; control 22) Age, years: adenosine (mean 67.3 ± 15.6), control (mean 66.3 ± 11.7) Sex: 65% men Inclusion: suboptimal reperfusion (< 70% STRes with persistent ST‐elevation > 2 mV in at least 1 anterior lead and > 1 mV in a non‐anterior lead) more than 10 minutes after last balloon inflation Exclusion: haemodynamic instability; prior myocardial infarction; an ECG unsuitable for calculation of STRes (left bundle branch block, paced or severe disturbed rhythm); history of obstructive pulmonary disease Withdrawals or losses to follow‐up: 1 from adenosine group
Interventions	Adenosine Dose: 60 mg in 10 mL saline infused for 5 to 10 minutes Route of administration: intracoronary Timing: during PPCI, 10 minutes after balloon inflation Control Dose: 10 mL saline infused for 5 to 10 minutes Route of administration: intracoronary Timing: during PPCI, 10 minutes after balloon inflation
Outcomes	Death (2 non‐cardiac deaths and 1 cardiac death at 12‐month follow‐up) TIMI (given by mean, adenosine 2.6, placebo 2.6) MBG (given by mean) Follow‐up: hospital stay and 12 months (by telephone interview)
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Randomised, no further description
Allocation concealment (selection bias)	Unclear risk	Not reported
Blinding (performance bias and detection bias) All outcomes	Low risk	"Patient and operators were blinded for the study medication"
Incomplete outcome data (attrition bias) All outcomes	High risk	1 lost to follow‐up in adenosine group, but no intention‐to‐treat analysis was reported and no reason was given
Selective reporting (reporting bias)	Low risk	All outcomes stated in Methods are reported in Results
Other bias	Unclear risk	Not obvious

Taniyama 1997.

Methods	Date of publication: 1997 Design: not reported Allocation: randomised Country: Japan
Participants	Diagnosis: acute AMI for PPCI Number: 40 (verapamil 20; control 20) Age, years: verapamil (mean 61 ± 11), control (mean 67 ± 14) Sex: 78% men Inclusion: diagnosis of a first AMI; single‐vessel disease; percutaneous transluminal coronary angioplasty of a totally occluded infarct‐related artery; TIMI flow grade 0 within 12 hours of symptom onset; residual diameter stenosis < 50%; no ischaemic event during follow‐up; adequate quality of echocardiogram Exclusion: severe congestive heart failure Withdrawals or losses to follow‐up: not reported
Interventions	Verapamil Dose: 0.5 mg in 10 mL saline Route of administration: intracoronary and orally until follow‐up Timing: shortly after PTCA Control: do not give anything
Outcomes	TIMI flow grade 3 after PTCA (verapamil 70%, control 70%) Follow‐up: mean of 25 days after AMI
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Randomised, no further description
Allocation concealment (selection bias)	Unclear risk	Not stated
Blinding (performance bias and detection bias) All outcomes	Unclear risk	No clear statement
Incomplete outcome data (attrition bias) All outcomes	Low risk	No incomplete outcome data
Selective reporting (reporting bias)	Low risk	All outcomes stated in Methods are reported in Results
Other bias	Low risk	Not obvious

Wang 2008.

Methods	Date of publication: 2008 Design: not reported Allocation: randomised Country: China
Participants	Diagnosis: AMI for PPCI Number: 28 (adenosine 14; control 14) Age, years: adenosine (mean 48.6 ± 13.7); control (mean 46.2 ± 13.1) Sex: 82% men Inclusion: chest pain for at least 30 minutes; time from onset of symptoms of 12 hours before hospital admission; an ECG showing ST‐segment elevation of 0.1 mV in 2 or more leads; cardiac marker increased to twice the upper limit; not received fibrinolysis at emergency department; HR ≥ 50/min, BP ≥ 90/60 mmHg Exclusion: cardiogenic shock; Killip class ≥ 3; history of asthma; HR < 50/min, TIMI flow 3 Withdrawals or losses to follow‐up: not reported
Interventions	Adenosine Dose: 500 μg in 10 mL normal saline Route of administration: intracoronary Timing: during PPCI, after balloon inflation Control Dose: 10 mL normal saline Route of administration: intracoronary Timing: during PPCI, after balloon inflation
Outcomes	TIMI (RR 0.50, 95% Cl 0.11 to 2.30) Bradycardia Follow‐up: during hospitalisation
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Randomised, no further description
Allocation concealment (selection bias)	Unclear risk	Not stated
Blinding (performance bias and detection bias) All outcomes	Unclear risk	No description
Incomplete outcome data (attrition bias) All outcomes	Low risk	No incomplete outcome data
Selective reporting (reporting bias)	Low risk	All outcomes stated in Methods are reported in Results
Other bias	Low risk	Not obvious

Zhang 2012.

Methods	Date of publication: 2012 Design: not reported Allocation: randomised Country: China
Participants	Diagnosis: AMI for PPCI Number: 59 (adenosine 32; control 27) Age, years: adenosine (mean 59.4 ± 10.2); control (mean 65.0 ± 12.2) Sex: 81% men Inclusion: had typical chest pain presenting within 12 hours of onset, with ST‐segment elevation in at least 2 contiguous leads of > 0.2 mV in precordial leads, > 0.1 mV in limb leads or new left bundle branch block (LBBB); candidates for primary PCI treatment Exclusion: thrombolytic treatment before PCI treatment; previous myocardial infarction; history of coronary artery bypass graft (CABG) or PCI; clinical evidence of significant reactive airway disease (e.g. asthma); advanced cancer, or end‐stage disease Withdrawals or losses to follow‐up: not reported
Interventions	Adenosine: 70 μg/kg/min for 3 hours Route of administration: intravenous Timing: immediately after the guide wire crossed the lesion and was positioned at the distal portion of the infarct‐related artery (IRA) Control Normal saline: 1.0 mL/min for 3 hours Route of administration: intravenous Timing: immediately after the guide wire crossed the lesion and was positioned at the distal portion of the infarct‐related artery (IRA)
Outcomes	Death Non‐fatal MI TIMI (RR 0.10, 95% Cl 0.01 to 0.76) Side effects: bradycardia, AV block Follow‐up: 6 months
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Randomised, no further description
Allocation concealment (selection bias)	Unclear risk	Not stated
Blinding (performance bias and detection bias) All outcomes	Unclear risk	No description
Incomplete outcome data (attrition bias) All outcomes	Low risk	No incomplete outcome data
Selective reporting (reporting bias)	Low risk	All outcomes stated in Methods are reported in Results
Other bias	Low risk	Not obvious

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Babbitt 1988	Not an RCT
Dani 2009	Not an RCT. Intervention groups did not meet inclusion criteria
De Luca 2012	Participants received elective PCI, which did not meet inclusion criteria
Dillon 2001	Not an RCT
Fischell 1998	Not an RCT. Participants received saphenous vein graft percutaneous coronary intervention
Forman 2008	Not an RCT
Garratt 1998	Not an RCT. Intervention was a combination of adenosine plus lidocaine, with no control group
Granger 1997	Not an RCT
Hang 2005	Single‐centre, non‐randomised, prospective study with a retrospective control group
Harding 2006	Not an RCT
Heidland 2000	Participants were patients with stable angina
Huang 2009	Abstract only, cannot contact study authors
Ibrahim 2002	Not an RCT
Jing 2010	Abstract only, cannot contact study authors
Kelly 2005	Not an RCT
Kim 2009	Participants had stable or unstable angina and received non‐urgent PCI, which did not meet inclusion criteria
Kloner 2006	Participants received both PPCI and thrombolysis as reperfusion therapy
Kopecky 2003	Intervention was adenosine agonist (AMP579), which did not meet inclusion criteria
Lee 2007	Participants underwent non‐urgent PCI, which did not meet inclusion criteria
Lim 2004	Intervention group got adenosine plus nicorandil; control group was given adenosine alone
Masaya 2004	Retrospective study, not an RCT
Matsuo 2007	Participants were not patients with AMI
Micari 2004	Abstract only, cannot contact study authors
Michaels 2002	Participants received saphenous vein graft percutaneous coronary intervention
Parikh 2004	Not an RCT; intervention group did not meet inclusion criteria
Parikh 2007	Not an RCT
Rezkalla 2010	Not an RCT. Retrospective study
Ross 2005	Participants received both PPCI and thrombolysis as reperfusion therapy
Sakuma 2010	Study was a prospective, non‐randomised, open‐label trial
Tian 2008	Participants underwent non‐urgent PCI, which did not meet inclusion criteria
Tsao 2009	Complex coronary artery disease. Not an RCT
Vijayalakshmi 2006	Participants were not only STEMI but also had non‐STEMI and unstable angina (acute coronary syndrome), so did not meet inclusion criteria
Werner 2002	Not an RCT. AMI participants received not only direct PTCA but also rescue PTCA
Wilkinson 2010	Not an RCT. Participants received saphenous vein graft percutaneous coronary intervention. Conference proceeding report
Zhen‐Guo 2010	Abstract only, cannot contact study authors

Characteristics of ongoing studies [ordered by study ID]

Garcia‐Dorado 2008.

Trial name or title	Myocardial protection during reperfusion in patients with acute coronary syndrome with ST‐segment elevation submitted to primary angioplasty: effect of intracoronary adenosine on infarct size and ventricular remodeling
Methods	RCT
Participants	STEMI
Interventions	Intracoronary adenosine
Outcomes	Infarct size measured by MRI
Starting date	October 2008
Contact information
Notes	http://clinicaltrials.gov/show/NCT00781404

Gershlick 2011.

Trial name or title	Randomized controlled trial comparing intracoronary administration of adenosine or sodium nitroprusside to control for attenuation of microvascular obstruction during primary percutaneous coronary intervention
Methods	RCT
Participants	STEMI undergoing primary percutaneous coronary intervention
Interventions	Intracoronary adenosine or sodium nitroprusside to control
Outcomes	Cardiac magnetic resonance measured infarct size, degree of ST resolution, incidence before and after angiographic procedure, true no‐reflow, overall major cardiac events
Starting date	27 April 2011
Contact information
Notes	https://www.clinicaltrialsregister.eu/ctr‐search/search?query=eudract_number:2010‐023211‐34

Niccoli 2009b.

Trial name or title	Randomized evaluation of intracoronary nitroprusside vs adenosine after thrombus aspiration during primary percutaneous coronary intervention for the prevention of no‐reflow in acute myocardial infarction: the REOPEN‐AMI study protocol
Methods	RCT
Participants	AMI patients undergoing primary or rescue percutaneous coronary intervention and thrombus aspiration
Interventions	Receive intracoronary nitroprusside, adenosine or placebo
Outcomes	ST‐segment resolution, TIMI < 3, major cardiac events
Starting date	December 2007
Contact information
Notes	NCT00590070

Contributions of authors

Review authors contributed to the updated review by:

• reviewing selected literature (LL, SQ);

• appraising the papers (LL, NTS);

• extracting the data (SQ, NTS);

• entering data into Review Manager 5 (SQ, NTS);

• conducting the analyses (SYH, SQ);

• interpreting the analyses (SYH, LL); and

• drafting the final updated review (LL, NTS, SQ).

Sources of support

Internal sources

• Guangxi Medical University, China.

External sources

• No sources of support supplied

Declarations of interest

None known.

New search for studies and content updated (conclusions changed)