Cardioprotective role of ischemic postconditioning in acute myocardial infarction: A systematic review and meta-analysis Abdur Rahman Khan, MD, a Aref A. Binabdulhak, MD, b Yaseen Alastal, MD, a Sobia Khan, MBBS, a Bridget M. Faricy-Beredo, MLIS, c Faraz Khan Luni, MD, a Wade M. Lee, MLS, c Sadik Khuder, PhD, a and Jodi Tinkel, MD a Toledo, OH and Kansas, MO

Background Evidence suggests that ischemic postconditioning (IPoC) may reduce the extent of reperfusion injury. We performed a meta-analysis of randomized controlled trials, which compared the role of IPoC during primary percutaneous coronary intervention (PCI) to PCI alone (control group) in ST-segment elevation myocardial infarction. Methods

Several databases were searched, which yielded 19 studies. The outcomes of interest were measures of myocardial damage (serum cardiac enzymes and infarct size by imaging) and left ventricular function (left ventricular ejection fraction and wall motion score index). Mean difference (MD) and standardized mean difference (SMD) were used to assess the treatment effect. An inverse variance method was used to pool data into a random-effects model.

Results Ischemic postconditioning demonstrated a decrease in serum cardiac enzymes (SMD −0.48, 95% CI −0.92 to −0.05, I 2 = 92%), reduction in infarct size by imaging (SMD −0.30, 95% CI −0.58 to −0.01, I 2 = 80%), wall motion score index (MD −0.19, 95% CI −0.29 to −0.09, I 2 = 44%), and showed improvement in left ventricular ejection fraction (IPoC 52 ± 0.4, control 49.7 ± 0.4) (MD 2.78, 95% CI 0.66-4.91, I 2 = 69%). All included studies were limited by high risk of performance and publication bias. Conclusions Ischemic postconditioning during PCI in ST-segment elevation myocardial infarction appears to be superior to PCI alone in reduction of both myocardial injury or damage and improvement in global and regional left ventricular function. The effect seems to be more pronounced when a greater myocardial area is at risk. Given the limitations of the current available evidence, additional data from large randomized controlled trials are warranted. (Am Heart J 2014;168:512-521.e4.)

Approximately 525,000 acute myocardial infarction (AMI) events along with 125,464 related deaths occur annually in the United States alone. 1 Early myocardial reperfusion with the use of primary percutaneous coronary intervention (PCI) has been the preferred treatment modality in ST-segment elevation myocardial infarction (STEMI). 2 Despite advances in revascularization strategies, the mortality from AMI has remained relatively constant in recent years. 3 Moreover, there is an increased incidence of long-term morbidity including ischemia-associated heart failure. 4 Myocardial reperfusion injury has been postulated to partially explain these outcomes. 5 From the aDepartment of Internal Medicine, University of Toledo Medical Center, Toledo, OH, bDepartment of Internal Medicine, University of Missouri – Kansas City, Kansas, MO, and cMulford Health Science Library – University of Toledo, Toledo, OH. Submitted January 22, 2014; accepted June 15, 2014. Reprint requests: Jodi Tinkel, MD, Division of Cardiovascular Medicine, Department of Medicine, University of Toledo Medical Center, 3000 Arlington Avenue, Toledo, OH 43614. E-mail: [email protected] 0002-8703 © 2014, Mosby, Inc. All rights reserved. http://dx.doi.org/10.1016/j.ahj.2014.06.021

The seminal work of Zhao et al 6 has generated interest in ischemic postconditioning (IPoC) to decrease the extent of myocardial damage. The proposed beneficial effect of IPoC is thought to be secondary to its effect on various cellular mechanisms: the kinase pathway, survival activating factor enhancement pathway, Janus signal transducer and activator system, reduction in formation of reactive oxygen or nitrogen species, and attenuation of calcium overload. 7 All of these mediators decrease lethal reperfusion injury and modulate apoptotic cardiomyocyte death. 7 There have been several studies that evaluated the role of IPoC in AMI. 8-20 Meta-analyses of these studies have shown a cardioprotective effect of IPoC in reperfusion injury. 21,22 However, there have been several subsequent studies published, some with conflicting results, failing to show benefit or even demonstrating harm. 23-30 Given this ongoing controversy, we performed a contemporary systematic review and meta-analysis that expands on the findings of the previous meta-analyses by inclusion of several new studies.

American Heart Journal Volume 168, Number 4

Methods Data sources and search strategy The systematic review was carried out in accordance with the preferred reporting items for systematic reviews and meta-analyses (PRISMA) guidelines. 31 The search strategy and subsequent literature search was performed by experienced medical reference librarians (B.F., W.L.). The search strategies were developed in PubMed and translated to match the subject headings and keywords for Embase, Cochrane Central Register of Controlled Trials, ISI Web of Science, and Scopus from last 10 years through May 5, 2014. The following MeSH, Emtree, and keyword search terms were used in combination: ischemic postconditioning, staccato reperfusion, cardioprotection, myocardial infarction, myocardial injury, coronary angioplasty, primary percutaneous intervention, controlled trials, intervention study, and randomized controlled trials (RCTs). The search accounted for plurals and variations in spelling with the use of appropriate wildcards. To identify further articles, we hand searched related citations in review articles and commentaries. All results were downloaded into EndNote (Thompson ISI ResearchSoft, Philadelphia, USA), and duplicate citations were identified and removed. Study selection Two authors (S.K., Y.A.) independently assessed the eligibility of identified studies. The studies that were evaluated were RCTs that focused on the role of IPoC in STEMI. Published abstracts or unpublished data were not included, as there is a discrepancy between full publications and unpublished data or published abstracts. 32 Data extraction Three reviewers (A.B., F.K.L., Y.A.) independently extracted data on population under study, patient characteristics, postconditioning protocols used, and relevant outcomes measured. We did not specify a priori outcome definitions, and they were accepted as defined in individual studies. The outcomes measured were markers of cardiac injury or damage (biomarker release and imaging) and indices of global and regional left ventricular function. Cardiac injury or damage was measured in terms of level of biomarker release (cardiac enzymes) and infarct size (IS) by imaging. Cardiac enzymes included serum levels or area under the curve of troponin, creatine kinase isoenzyme MB fraction (CK-MB), or creatine kinase (CK). Infarct size was determined by cardiac magnetic resonance (CMR) or single photon emission computed tomography (SPECT). Global and regional ventricular functions were determined by left ventricular ejection fraction (LVEF) and wall motion score index (WMSI), respectively. Data synthesis and statistical analysis Continuous data were reported either as mean (SD) or median (interquartile range). If the data were reported as

Khan et al 513

median, mean and SD were estimated. 33 For continuous data, mean difference (MD) was calculated where same scale was used to measure relevant outcomes (LVEF and WMSI), whereas standardized mean difference (SMD) was considered where different units of measurements were used (IS by cardiac enzyme and imaging). A randomeffects model was used to pool data, and corresponding forest plots were constructed. Cochran Q test was used to assess heterogeneity among studies and was complemented by the I 2 statistic. 34 Heterogeneity was investigated by means of a subgroup analysis. Subgroups were divided based on geographic regions where the studies were conducted (Asia vs North America/Europe), individual cardiac enzymes, imaging modalities used to measure IS or LVEF, and involvement of the infarct-related artery (IRA). When grouped based on the IRA, studies were stratified depending upon left anterior descending (LAD) occlusion being N50% or b50%. Sensitivity analysis was carried out using a fixedeffect model meta-analysis. Publication bias was assessed by funnel plot, and Begg and Mazumdar test was done to assess funnel plot asymmetry and publication bias. 35

Quality assessment Two reviewers (A.B., S.K.) independently assessed the methodological quality of selected studies using the Cochrane risk of bias tool. This scale is used to explore adequacy of sequence generation, allocation sequence concealment, blinding of participants and caregivers, blinding for outcome assessment, incomplete outcome, selective outcome reporting, and other potential bias. 36 Any disagreements between reviewers in study inclusion, data extraction, and quality assessment that could not be resolved by consensus were resolved by a third reviewer (J.T.). All analyses were conducted using the statistical software Review Manager (RevMan) Version 5.2. Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2012 and (version 9.2, SAS Institute Inc. Cary, NC, USA). The authors are solely responsible for the design and conduct of this study and its final contents. No extramural funding was used to support this work.

Results Identification of studies The literature search identified 506 publications, of which 19 studies were eligible for analysis (Figure 1).8-14,16-19,23-25,27-30 Articles reporting outcomes from the same study were included once in the analysis. 9,10,14,15 There was excellent agreement for the inclusion of the studies, data abstraction, and quality assessment between the reviewers. Study characteristics Table summarizes the characteristics of the included studies. A total of 19 studies comprising 1,844 participants (ranging from 38 to 700 in individual trials) were

American Heart Journal October 2014

514 Khan et al

Figure 1

Flow diagram of eligible studies.

included in the analysis. There were 920 patients who underwent IPoC during PCI and 924 who underwent PCI alone. The risk of bias from all included studies was determined to be high due to risk of performance bias as none of the studies were blinded (Figure 2). The characteristics of the included patients are summarized in online Appendix Supplementary Table I. The IPoC protocol (cycles × ischemia/reperfusion in seconds) varied between studies being 2 × 90/180 in 1 study, 13 3 to 4 × 30/30 in 8 studies, 9,11,14,18,20,28-30 and 4 × 60/60 in 10 studies. 8,12,16,17,19,23-27 The mean follow-up in the trials was 90 days (3-455 days). The relevant outcomes of interest in all studies were markers of cardiac injury or damage and global and regional left

ventricular function (online Appendix Supplementary Tables II and III). In the included studies, troponin levels were measured in 5 studies, 12,17,23,25,29 CK-MB levels in 8 studies, 9,16-18,20,25,29,30 and CK in 10 studies 8,9,11-13,18,20,24,25,28 (online Appendix Supplementary Table II). The preference of inclusion of specific cardiac enzymes in the analysis was troponin levels followed by CK-MB and CK. Single photon emission computed tomography 11,12,16 and CMR 14,17,23-28 were the imaging modalities used to assess IS. There were also some differences in methods used to report IS; 3 studies reported IS as percentage of the area at risk (AAR), 17,27,28 4 studies reported IS as a percentage of the left ventricular

American Heart Journal Volume 168, Number 4

Khan et al 515

Table. Characteristics of included studies Study, country

Study design, center

Staat, France

RCT, multicenter

Ma, China Yang, China

RCT, single RCT, single

Thibault, France

RCT, 3 centers

Laskey, United States RCT, NR Lønborg, Denmark

RCT, single

Xue, China

RCT, single

Sörensson, Sweden

RCT, single (PROBE) Garcia, United States RCT, multicenter

Zhao, China

RCT, single

Liu, China

RCT, single

Tarantini, Italy

RCT, single (PROBE)

Thuny, France

RCT, NR

Freixa, Spain Sörensson, Sweden

RCT, single (PROBE) RCT, single

Mewton, France

RCT, multicenter

Dwyer, Canada

RCT, single

Dong, China

RCT, single

Hahn, Korea

RCT, multicenter (PROBE)

Inclusion criteria

Exclusion criteria

AMI, PCI, age N18 y, within 6 h, Previous MI, cardiac arrest, LAD/RCA, TIMI 0 cardiogenic shock, postintervention TIMI b2-3 STEMI, PCI in 12 h Prior CABG/MI, history of PCI STEMI, PCI; ST-segment elevation Previous MI, cardiogenic shock; TIMI ≥1; N0.1 mV, no collaterals left main occlusion, GpIIb-IIIa STEMI, PCI, LAD/RCA, Previous MI, cardiac arrest, cardiogenic age N18 yr, 6 hr, TIMI 0 shock, preinfarction angina, collaterals Anterior STEMI, PCI, chest pain Previous MI/CABG, thrombolysis, 30 m-6 h, TIMI 0-1, no collateral age N90 y, hemodynamic instability STEMI, PCI, age N18 y, Previous CABG/MI, TIMI N1, b12 h from onset postreperfusion TIMI b2, cardiogenic shock, LBBB, renal/liver failure STEMI, PCI, ST-segment Pain N12 h, preinfarction angina, elevation N0.1 mV prior MI, cardiac arrest. cardiogenic shock, collaterals, TIMI ≥0 STEMI, PCI; chest pain Prior CABG/MI/cardiogenic shock/arrest, 30 min-6 hr, LBBB, TIMI 0 renal failure, persistent Afib STEMI, PCI, 12 h from pain, Prior CABG/MI, collateral TIMI 0 STEMI; age N18 y, b12 h Previous MI, cardiac arrest, cardiogenic from onset, TIMI 0 shock, preinfarction angina, collaterals STEMI, PCI, 12 h from pain, NSTEMI, fibrinolysis before PCI, TIMI 0, no collaterals prior CABG/MI, CHF, NYHA II STEMI; age N18 y, b6 h, Prior CABG/MI, cardiogenic shock, TIMI ≤1, PPCI + abciximab collaterals, arrhythmia, abciximab/CMR contraindicated, renal/liver failure STEMI, PCI, age ≥18 y, Prior MI/cardiac arrest/cardiogenic within 12 h, TIMI 0-1, shock/VF, angina, CMR contraindicated no collaterals STEMI, PCI, age N18 y, Prior MI, TIMI 2-3, cardiogenic shock, within 12 h renal insufficiency, CMR contraindicated STEMI, PCI; age N18 y, Prior CABG/MI, cardiogenic shock, pain 30 m-6 h, TIMI 0 CMR contraindicated, Afib STEMI, PCI, age N18 y, Prior MI, cardiac arrest, cardiogenic shock, 12 h from pain, TIMI 0-1 preinfarction angina, collateral, CMR contraindicated STEMI, PCI, age ≥18 y, Cardiogenic shock, cardiac arrest, within 6 h, TIMI 0 prior MI/CABG Prolonged chest pain, N12 h, Prior CABG/MI/cardiogenic STEMI, PCI within 12 h shock/arrest STEMI, PCI, pain Cardiogenic shock, LBBB, left main lesion, for b12 h, TIMI 0-1 rescue PCI/facilitated PCI, life expectancy b1 y

Outcome

Follow-up

CK, blush grade

3d

CK-MB, WMSI, CTFC CK, STR, IS, LVEF

8 wk 7d

IS

6 m, 1 y

STR, CFVR

None

IS, STR

3 m, 15 m

CK-MB, IS, LVEF, STR

7d

CK-MB, LVEF, IS, troponin IS, MPG, LVEF

6-9 d

LVEF, WMSI

1 wk, 6 m

CK, CK-MB, WMSI, LVEF IS

8 wk

IS, peak CK

48-72 h

IS, LVEF

1 wk, 6 ms

IS, LVEF

3 and 12 m

IS, MVO

4d

IS

3d

3 m, 3.4 y

30 ± 10 d

CK-MB, LVEF, troponin 7 d STR, blush grade

30 d

PoC protocol: duration of balloon inflation × number of inflation. Abbreviations: RCA, right coronary artery; TIMI, thrombolysis in myocardial infarction; CTFC, corrected TIMI frame count; GpIIb-IIIa, platelet glycoprotein IIb/IIIa complex; STR, STsegment resolution; CFVR, coronary flow velocity reserve; LBBB, left bundle branch block; Afib, atrial fibrillation; MPG, myocardial perfusion grade; CHF, chronic heart failure; NYHA II, New York Heart Association class II; PPCI, primary percutaneous coronary intervention; VF, ventricular fibrillation; MVO, microvascular obstruction.

mass, 14,23,25,26 whereas 1 study reported it as gram 24 (online Appendix Supplementary Table II). Global left ventricular function as determined by LVEF was measured in 8 studies by echocardiography, 11-13,16,19,20,23,29 in 4 studies by CMR, 15,17,26,28 in 1 study by both echocardiography and CMR, 25 and in 1 study by echocardiography and CMR in each group, respectively. 18 Regional left ventricular function was measured by WMSI. 9,12,16,19,20 A score of 1 was considered normal with higher scores suggestive of abnormal regional systolic function (online

Appendix Supplementary Table II). Left anterior descending was the culprit artery in N50% of the patients in 9 studies, 9,11-13,16,19,20,24,27 whereas in the rest of the studies, LAD involvement was b50%.

Meta-analysis of markers of cardiac injury or damage Cardiac enzymes. Ischemic postconditioning showed a decrease in the level of cardiac enzymes after AMI (SMD −0.48, 95% CI −0.92 to 0.05, P = .03). There was substantial between-study heterogeneity (Cochran Q test,

516 Khan et al

Figure 2

American Heart Journal October 2014

Subgroup analysis based on IRA showed that, not only the decrease in cardiac enzymes became more significant, but the heterogeneity also dropped with LAD involvement N50% (SMD −0.81, 95% CI −1.03 to −0.60, I 2 = 0%, P = .00001) as compared with LAD b50% (SMD −0.19, 95% CI −0.85 to 0.47, I 2 = 95%, P = .57) (Figure 3).There was no significant change in heterogeneity when the studies were grouped based on geographic region or individual cardiac enzymes. IS by imaging. The pooled outcome of studies suggested a reduction in IS as measured by imaging (SMD −0.44, 95% CI −0.82 to −0.06, P = .02). There was substantial between-study heterogeneity (Cochran Q test, P b .00001, I 2 = 83%) (Figure 4).There was some visual evidence of funnel asymmetry, and Begg test confirmed the publication bias (P = .002) (online Appendix Supplementary Figure 2). A fixed-effect model also revealed a similar decrease in IS (SMD −0.30, 95% CI −0.45 to 0.15, P = .0001) Subgroup analysis based on IRA showed that, not only the decrease in IS became more significant, but the heterogeneity also dropped with LAD involvement N50% (SMD −1.05, 95% CI −1.36 to −0.74, I 2 = 14%, P = .00001) as compared with LAD b50% (SMD 0.01, 95% CI −0.34 to 0.35, I 2 = 72%, P = .97) (Figure 4). There was no significant change in heterogeneity when the studies were grouped based on geographic region.

Forest plot showing the WMSI, expressed as MD using a randomeffects model. Stratified based on LAD artery involvement.

P b .00001, I 2 = 92%) (Figure 3). There was some suggestion of funnel asymmetry; publication bias was confirmed by Begg test (P = .03) (online Appendix Supplementary Figure 1). A fixed-effect model also revealed a similar decrease in biomarker release (SMD −0.18, 95% CI −0.29 to −0.08, P = .0006).

Meta-analysis of markers of cardiac function Left ventricular ejection fraction. The analysis showed a mild improvement in LVEF (MD 4.34, 95% CI 1.21-7.47, P = .007) with between-study heterogeneity (Cochran Q test, P b .0001, I 2 = 75%). There was no visual evidence of funnel asymmetry, and Begg test did not show publication bias (P = .24) (online Appendix Supplementary Figure 3). A fixed-effect model also revealed a similar improvement in LVEF (MD 3.26, 95% CI 1.78-4.74). Subgroup analysis based on IRA showed that the improvement in LVEF became more significant, and the heterogeneity dropped with LAD involvement N50% (MD 5.89, 95% CI 3.61-8.16, I 2 = 0%, P = .00001) as compared with LAD b50% (MD 2.57, 95% CI −2.51 to 7.66, I 2 = 84%, P = .32) (Figure 5). There was no significant change in heterogeneity when the studies were grouped based on geographic region. Heterogeneity dropped to 0% when CMR was used with no significant change with echocardiography as an imaging modality. Wall motion score index. Our analysis showed that IPoC reduced WMSI as compared with PCI alone (MD −0.19, 95% CI −0.29 to −0.09, P = .00001, I 2 = 44%) (Figure 6). There was no publication bias (online Appendix Supplementary Figure 4) (P = .33).

American Heart Journal Volume 168, Number 4

Khan et al 517

Figure 3

Risk of bias in individual studies.

Figure 4

Forest plot showing cardiac enzymes levels, expressed as SMD using a random-effects model. Stratified based on LAD artery involvement.

518 Khan et al

American Heart Journal October 2014

Figure 5

Forest plot showing the IS by imaging, expressed as SMD using a random-effects model. Stratified based on LAD artery involvement.

Discussion In our meta-analysis, we found that IPoC showed a beneficial effect in reduction of myocardial injury or damage and improvement of left ventricular function when compared with PCI alone. This cardioprotection was more apparent when the IRA was predominantly LAD as compared with non-LAD. Our results suggest that myocardial salvage is better in the anterior location compared with the inferior location because of a greater myocardial area being at risk. The lack of substantial effect of IPoC in non-LAD coronary occlusion was most likely due to a smaller AAR. Moreover, the sample size of the studies may have been small to detect minor beneficial effects with non-LAD coronary occlusions. This underlines the importance of adequate and uniform assessment of myocardial AAR. The assessment of AAR is affected by the vessel involved, presence of coronary collaterals, and imaging modality used. A pooled analysis of RCTs has shown that involvement of the LAD is one of the strongest predictors of IS. 37 It has been shown that a protective effect of IPoC is apparent when the AAR is N30%. 17 Although the presence of collaterals and incomplete coronary occlusion has an important bearing on the AAR, all the included studies did not have patients with any collateral circulation. The timing of imaging and the modality used is important in temporal relation to reperfusion as changes in myocardial water content may affect the estimate of AAR. 38,39 Although CMR has

been shown to be superior to SPECT to measure IS, it overestimates the AAR postreperfusion due to the myocardial edema. 38 Thus, there should be uniformity in assessment of the AAR and subsequent myocardial injury by the clinical imaging method used. Although our analysis suggested a beneficial effect of IPoC, the cumulative evidence in favor of IPoC was negatively affected by the variation in results, heterogeneity in the studies, high risk of performance bias, and the presence of publication bias. The variation in results may be partly explained by a predominant non-LAD involvement in IRA. Pooling the results with predominant LAD occlusion of the IRA showed both consistency in results and resolution of the heterogeneity. The included studies also differed in patient’s baseline characteristics, coexisting diseases, use of concomitant medications, ischemia time, postconditioning protocols used, and definition of a successful outcome. These confounding factors may partly account for the heterogeneity encountered and may partially explain the variation in results. The effect of age, gender, various comorbidities, and use of concomitant medications can potentially affect the cardioprotective benefits of postconditioning. 40 One approach would be to pool patient-level data, which would help to assess variation in treatment effect in patients with variation in individual characteristics.

Strengths and limitations As compared with previous reviews 21,22 our analysis involves a comprehensive literature search with inclusion

American Heart Journal Volume 168, Number 4

Khan et al 519

Figure 6

Forest plot showing the LVEF, expressed as MD using a random-effects model. Stratified based on LAD artery involvement.

of the largest number of relevant studies. We also have examined the potential effect of myocardial AAR to explain the observed association. The results of our meta-analysis are weakened by limitations inherent to the studies included in the analysis. The included studies were limited by small sample size, marked between-study heterogeneity, and high risk of performance bias that affects the interpretation of the results of our analysis.

Future directions Ischemic postconditioning is an exciting, cost-effective and easily reproducible method to decrease the magnitude of reperfusion injury. There are several questions that need to be addressed before IPoC can be used routinely in a large-scale clinical setting. Patient selection would be crucial as most of the patients with AMI would be associated with confounders in the form of comorbidities and concomitant medications in the real-world scenario 41 as would be the determination of AAR. Risk prediction modeling could be used to identify the subgroups that will benefit most from cardioprotection. Moreover, beside patient selection, the role of different reperfusion strategies and the timing of the intervention need to be addressed as recent studies have suggested little or no benefit with gentle reperfusion 42 or a delayed reperfusion approach. 43 The efficacy of IPoC needs to be determined further with the increasing use of thrombus aspiration and mechanical thrombectomy. The long-term outcomes of IPoC when used in routine clinical practice are still unclear despite evidence in favor of reduction of IS. A retrospective analysis with multiple balloon inflations done during PCI as a “real world” analog for IPoC did show a reduction in IS as measured by enzymatic release but failed to translate into long-term beneficial outcomes in terms of mortality or major adverse cardiovascular outcomes. 44 These factors should be kept in mind while designing future studies to evaluate the role of IPoC and its long-term benefits as the target patient population will have associated comorbidities, will use concomitant medications, and will have different interventional proce-

dures performed at different times, which may limit the clinical benefit observed in the real world. 41 Remote ischemic preconditioning with cycles of transient limb ischemia with a blood pressure cuff is seen as an alternative to IPoC with a recent study showing benefit in reduction of IS by enzymatic release. 45 Remote preconditioning may be an option as it is not only easy to administer but also affects the same final signaling pathway for cardioprotection. 46 It will also need further studies to support and confirm these findings.

Conclusions Ischemic postconditioning during PCI in STEMI appears to be superior to PCI alone in reduction of both myocardial injury or damage and improvement in global and regional left ventricular function. The effect seems to be more pronounced when a greater myocardial area is at risk secondary to LAD involvement. Given the limitations of the current available evidence, additional data from large RCTs are warranted.

References 1. Go AS, Mozaffarian D, Roger VL, et al. Heart disease and stroke statistics–2013 update: a report from the American Heart Association. Circulation 2013;127(1):e6-245. 2. O'Gara PT, Kushner FG, Ascheim DD, et al. 2013 ACCF/AHA guideline for the management of ST-elevation myocardial infarction: a report of the American College of Cardiology Foundation/ American Heart Association Task Force on Practice Guidelines. Circulation 2013;127(4):e362-425. 3. Menees DS, Peterson ED, Wang Y, et al. Door-to-balloon time and mortality among patients undergoing primary PCI. N Engl J Med 2013;369(10):901-9. 4. Fox KA, Carruthers KF, Dunbar DR, et al. Underestimated and under-recognized: the late consequences of acute coronary syndrome (GRACE UK-Belgian Study). Eur Heart J 2010;31(22):2755-64. 5. Yellon DM, Hausenloy DJ. Myocardial reperfusion injury. N Engl J Med 2007;357(11):1121-35. 6. Zhao ZQ, Corvera JS, Halkos ME, et al. Inhibition of myocardial injury by ischemic postconditioning during reperfusion: comparison

520 Khan et al

7. 8. 9.

10.

11.

12. 13.

14.

15.

16. 17.

18.

19.

20.

21.

22.

23.

24.

25.

with ischemic preconditioning. Am J Physiol Heart Circ Physiol 2003;285(2):H579-88. Heusch G. Cardioprotection: chances and challenges of its translation to the clinic. Lancet 2013;381(9861):166-75. Staat P, Rioufol G, Piot C, et al. Postconditioning the human heart. Circulation 2005;112(14):2143-8. Ma XJ, Zhang XH, Li CM, et al. Effect of postconditioning on coronary blood flow velocity and endothelial function in patients with acute myocardial infarction. Scand Cardiovasc J 2006;40(6):327-33. Ma XJ, Zhang XH, Liu M, et al. Effects of preconditioning and postconditioning on emergency precutaneous coronary intervention in patients with acute myocardial infarction. Natl Med J China 2007;87(2):114-7. Yang XC, Liu Y, Wang LF, et al. Reduction in myocardial infarct size by postconditioning in patients after percutaneous coronary intervention. J Invasive Cardiol 2007;19(10):424-30. Thibault H, Piot C, Staat P, et al. Long-term benefit of postconditioning. Circulation 2008;117(8):1037-44. Laskey WK, Yoon S, Calzada N, et al. Concordant improvements in coronary low reserve and ST-segment resolution during percutaneous coronary intervention for acute myocardial infarction: A benefit of postconditioning. Catheter Cardiovasc Interv 2008;72(2):212-20. Lonborg J, Kelbaek H, Vejlstrup N, et al. Cardioprotective effects of ischemic postconditioning in patients treated with primary percutaneous coronary intervention, evaluated by magnetic resonance. Circ Cardiovasc Interv 2010;3(1):34-41. Lonborg J, Holmvang L, Kelbaek H, et al. ST-Segment resolution and clinical outcome with ischemic postconditioning and comparison to magnetic resonance. Am Heart J 2010;160(6):1085-91. Xue F, Yang X, Zhang B, et al. Postconditioning the Human Heart in Percutaneous Coronary Intervention. Clin Cardiol 2010;33(7):439-44. Sörensson P, Saleh N, Bouvier F, et al. Effect of postconditioning on infarct size in patients with ST elevation myocardial infarction. Heart 2010;96(21):1710-5. Garcia S, Henry TD, Wang YL, et al. Long-term follow-up of patients undergoing postconditioning during ST-elevation myocardial infarction. J Cardiovasc Transl Res 2011;4(1):92-8. Zhao CM, Yang XJ, Yanga JH, et al. Effect of ischaemic postconditioning on recovery of left ventricular contractile function after acute myocardial infarction. J Int Med Res 2012;40(3):1082-8. Liu TK, Mishra AK, Ding FX. Protective effect of ischemia postconditioning on reperfusion injury in patients with ST-segment elevation acute myocardial infarction. Zhonghua Xin Xue Guan Bing Za Zhi 2011;39(1):35-9. Wei Y, Ruan L, Zhou G, et al. Local ischemic postconditioning during primary percutaneous coronary intervention: a meta-analysis. Cardiology 2012;123(4):225-33. Zhou C, Yao Y, Zheng Z, et al. Stenting technique, gender, and age are associated with cardioprotection by ischaemic postconditioning in primary coronary intervention: a systematic review of 10 randomized trials. Eur Heart J 2012;33(24):3070-7. Tarantini G, Favaretto E, Marra MP, et al. Postconditioning during coronary angioplasty in acute myocardial infarction: The POST-AMI trial. Int J Cardiol 2012;162(1):33-8. Thuny F, Lairez O, Roubille F, et al. Post-conditioning reduces infarct size and edema in patients with ST-segment elevation myocardial infarction. J Am Coll Cardiol 2012;59(24):2175-81. Freixa X, Bellera N, Ortiz-Pérez JT, et al. Ischaemic postconditioning revisited: Lack of effects on infarct size following primary percutaneous coronary intervention. Eur Heart J 2012;33(1):103-12.

American Heart Journal October 2014

26. Sorensson P, Ryden L, Saleh N, et al. Long-term impact of postconditioning on infarct size and left ventricular ejection fraction in patients with ST-elevation myocardial infarction. BMC Cardiovasc Disord 2013;13:22. http://www.biomedcentral.com/1471-2261/ 13/22. 27. Mewton N, Thibault H, Roubille F, et al. Postconditioning attenuates no-reflow in STEMI patients. Basic Res Cardiol 2013;108(6):383 http://dx.doi.org/10.1007/s00395-013-0383-8. 28. Dwyer NB, Mikami Y, Hilland D, et al. No cardioprotective benefit of ischemic postconditioning in patients with ST-segment elevation myocardial infarction. J Interv Cardiol 2013;26(5):482-90. 29. Dong M, Mu N, Guo F, et al. The beneficial effects of postconditioning on no-reflow phenomenon after percutaneous coronary intervention in patients with ST-elevation acute myocardial infarction. J Thromb Thrombolysis 2014 Aug;38(2):208-14 http://dx.doi.org/10.1007/ s11239. 30. Hahn JY, Song YB, Kim EK, et al. Ischemic postconditioning during primary percutaneous coronary intervention: the effects of postconditioning on myocardial reperfusion in patients with ST-segment elevation myocardial infarction (POST) randomized trial. Circulation 2013;128(17):1889-96. 31. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration. BMJ 2009;339:b2700 http://dx.doi.org/10.1136/bmj.b2700. 32. Scherer RW, Langenberg P, von Elm E. Full publication of results initially presented in abstracts. Cochrane Database Syst Rev 2007;2: MR000005. 33. Higgins JPT, Green S. The Cochrane Handbook for Systematic Reviews of Interventions. Version 5.1.0 (updated March 2011). The Cochrane Collaboration. 2011. 34. Higgins JP, Thompson SG, Deeks JJ, et al. Measuring inconsistency in meta-analyses. BMJ 2003;327(7414):557-60. 35. Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for publication bias. Biometrics 1994;50(4): 1088-101. 36. Higgins JP, Altman DG, Gotzsche PC, et al. The Cochrane Collaboration's tool for assessing risk of bias in randomised trials. BMJ 2011;343:d5928 http://dx.doi.org/10.1136/bmj.d5928. (Published 18 October 2011). 37. Stone GW, Dixon SR, Grines CL, et al. Predictors of infarct size after primary coronary angioplasty in acute myocardial infarction from pooled analysis from four contemporary trials. Am J Cardiol 2007;100(9):1370-5. 38. Mewton N, Rapacchi S, Augeul L, et al. Determination of the myocardial area at risk with pre- versus post-reperfusion imaging techniques in the pig model. Basic Res Cardiol 2011;106(6): 1247-57. 39. Wright J, Adriaenssens T, Dymarkowski S, et al. Quantification of myocardial area at risk with T2-weighted CMR: comparison with contrast-enhanced CMR and coronary angiography. J Am Coll Cardiol Img 2009;2(7):825-31. 40. Ferdinandy P, Schulz R, Baxter GF. Interaction of cardiovascular risk factors with myocardial ischemia/reperfusion injury, preconditioning, and postconditioning. Pharmacol Rev 2007;59(4):418-58. 41. Heusch G. Reduction of infarct size by ischaemic post-conditioning in humans: fact or fiction? Eur Heart J 2012;33(1):13-5. 42. Musiolik J, van Caster P, Skyschally A, et al. Reduction of infarct size by gentle reperfusion without activation of reperfusion injury salvage kinases in pigs. Cardiovasc Res 2010;85(1):110-7.

American Heart Journal Volume 168, Number 4

43. Roubille F, Mewton N, Elbaz M, et al. No post-conditioning in the human heart with thrombolysis in myocardial infarction flow 2-3 on admission. Eur Heart J 2014 Jul 1;35(25):1675-82 http://dx.doi. org/10.1093/eurheartj/ehu054. Epub 2014 Feb 28. 44. Yetgin T, Magro M, Manintveld OC, et al. Impact of multiple balloon inflations during primary percutaneous coronary intervention on infarct size and long-term clinical outcomes in ST-segment elevation myocardial infarction: real-world postconditioning. Basic Res Cardiol 2014;109(2):403 http://dx.doi.org/10.1007/s00395-014-0403-3.

Khan et al 521

45. Prunier F, Angoulvant D, Saint Etienne C, et al. The RIPOST-MI study, assessing remote ischemic perconditioning alone or in combination with local ischemic postconditioning in ST-segment elevation myocardial infarction. Basic Res Cardiol 2014;109 (2):400 http://dx.doi.org/10.1007/s00395-013-0400-y. Epub 2014 Jan 10. 46. Heusch G, Musiolik J, Kottenberg E, et al. STAT5 activation and cardioprotection by remote ischemic preconditioning in humans: short communication. Circ Res 2012;110(1):111-5.

American Heart Journal Volume 168, Number 4

Khan et al 521.e1

Appendix Table of Contents Supplemental Table 2. Outcomes Measured in the Included Studies (Markers of Myocardial Injury or Damage Supplemental Table 3. Outcomes Measured in the Included Studies (Markers of Myocardial Function) Supplemental Figure 1. Funnel plot to assess publication bias for association between ischemic postconditioning in acute myocardial infarction and cardiac enzymes Supplemental Figure 2. Funnel plot to assess publication bias for association between ischemic postconditioning in acute myocardial infarction and infarct size Supplemental Figure 3. Funnel plot to assess publication bias for association between ischemic postconditioning in acute myocardial infarction and subsequent left ventricular ejection fraction Supplemental Figure 4. Funnel plot to assess publication bias for association between ischemic postconditioning in acute myocardial infarction and wall motion score index

Page 3 Page 4 Page 5 Page 6 Page 7 Page 8

Supplementary Table I. Baseline Characteristics of the Included Patients in the Included Studies Culprit Artery (%) Study

Groups

Staat 9

IPoC Control IPoC Control IPoC Control IPoC Control IPoC Control IPoC Control IPoC Control IPoC Control IPoC Control IPoC Control IPoC Control IPoC Control IPoC Control IPoC Control IPoC Control IPoC Control IPoC Control IPoC Control IPoC Control

Ma

10,11

Yang

12

Thibault 13 Laskey 14 Lønborg

15,16

Xue 17 Sörensson 18 Garcia Zhao

19

20

Liu 21 Tarantini 25 Thuny 26 Freixa 27 Sörensson 28 Mewton 29 Dwyer

30

Dong

31

Hahn

32

n

Age (yr)

Male (%)

DM (%)

HPL (%)

HTN (%)

IPoC Protocol⁎

Ischemia time#

LAD

LCx

RCA

28 27 47 47 23 18 17 21 12 12 59 59 23 20 38 38 22 21 32 30 30 34 37 38 25 25 39 40 33 35 25 25 50 52 32 30 350 350

58 56 64 64 59 63 56 56 60 58 61 62 54 62 63 62 61 55 56 62 59 59 60 60 57 57 59 60 63 62 57 57 57 57 70 68 60 60

43 48 66 70 87 61 76 78 58 58 69 74 95 94 82 89 86 76 96 87 73 68 85 85 76 72 84 72 85 89 76 72 88 89 62 73 79 75

20 13 38 45 26 28 12 10 42 42 07 07 21 29 45 31 05 19 12 27 30 32 18 03 20 16 23 17 29 32 20 14 14 6 34 37 24 25

80 50 NR NR 61 56 52 49 58 75 46 41 16 24 71 55 73 71 31 17 NR NR 51 49 36 48 44 35 NR NR NR NR 36 29 NR NR 40 46

38 36 62 55 70 61 29 35 75 83 63 54 37 71 16 29 64 48 50 67 37 39 59 49 40 48 49 50 15 31 40 48 42 33 72 63 46 46

4 x 60/60

NR NR 6.6±2.5 h 7.1±2.5 h NR NR 283±82 297±104 228±43 222±54 241±149 255±196 4.1± 3.0 h 5.4± 3.7 h NR NR 4.5 h 4.4 h 309±201 404±191 310 324 212±85 194 ±80 289±31 215±20 326± 180 330± 211 NR NR 289± 31 215± 20 150± 70 170± 84 5.0 ±1.5 4.9 ± 1.1 196 195

21 22 49 53 65 61 56 52 100 100 44 39 42 59 37 37 36 24 53 63 53 58 41 44 56 56 51 45 33 37 56 56 50 46 56 43 47 45

23 17 09 06 08 19 10 03 23 10 NR NR 10 12 18 08 0 08

36 29 28 30 26 33 44 48 47 42 58 41 16 21 41 67 NR NR 37 29 41 49 44 36 39 47 57 60 0 08 38 40 37 47 43 44

3 x 30/30 3 x 30/30 4 x 60/60 2 x 90/180 4 x 30/30 4 x 60/60 4 x 60/60 4 x 30/30 4 x 60/60 3 x 30/30 4 x 60/60 4 x 60/60 4 x 60/60 4 x 60/60 4 x 60/60 4 x 30/30 3 x 30/30 4 x 60/60

09 03 44 36 12 14 6 10 11 11

IPoC Protocol: Cycles x I/D: Cycles x Inflation/Deflation; # Ischemia time: time in minutes and hours (h). Abbreviations: n, number; yr, years; DM, Diabetes Mellitus; HTN, Hypertension; HPL, Hyperlipidemia; IPoC,Ischemic Postconditioning; LAD, Left anterior descending artery; LCx, Left circumflex artery; RCA, Right coronary artery.

American Heart Journal October 2014

521.e2 Khan et al

Supplementary Table II. Outcomes Measured in the Included Studies (Markers of Myocardial Injury or Damage) Study

IPoC group

CARDIAC ENZYMES - CK Staat 9 Ma 10 Yang 12 Thibault 13 Laskey 14 Garcia 19 Liu Thuny 25⁎ Freixa 26 Dwyer 29

2831 1236 2699 22.7 1524 2182 1162 1695 3909 66.9

± ± ± ± ± ± ± ± ± ±

404 813 634 9.3 435 1717 548 1906 485 52.8

CARDIAC ENZYMES – CK-MB Ma 10 117 ± 76 Xue 17 247.7 ±118.3 4175 ± 2706 Sörensson 18⁎ Garcia 19 195 ± 33 Liu 165 ± 70 Freixa 26 251 ± 28.9 Dong 30 2159.9 ± 485.5 Hahn 31 232 ±172 CARDIAC ENZYMES - Troponin Thibault 13 13.0 ± Sörensson 18⁎ 165 ± Tarantini 24 112.4 ± 299 ± Freixa 26 Dong 30 154 ±

7.0 136.3 97.8 72 43.1

Control group

4234 1697 2229 37.9 1862 2444 1732 3505 3122 72.3

±722 IU/L ± 966 ± 255 ± 19.5 ± 561 ± 1928 ± 480 ± 1942 ± 379 ± 53.5

172 351.9 3890 242 280 195 2397.6 229

± 93 ±153.6 ± 2871 ± 40 ± 99 ± 17.6 ± 470.2 ± 204

24.6 147 89.2 148 197.5

± 20.6 ± 140 ± 66.0 ± 23.8 ± 32.5

INFARCT SIZE - SPECT Yang 12 Thibault 13 Xue 17

22.8 ± 6.7 11.8 ±10.3 13 ±11.2

31.3 ± 8.6 19.5 ± 13.3 24.2± 10.6

INFARCT SIZE - CMR Lønborg 15 Sörensson 18⁎ Tarantini 24 Thuny 25⁎ Freixa 26 (7 days) (6 months) Sörensson 27⁎ (1 week) (3 months) (12 months) Mewton 28 Dwyer 29 (1 week) (6 months)

14 47 20.2 13 27.5 21.8 9.9 8.8 7.6 18.0 13 09

17 ± 8 44 ± 19.3 14.3 ± 9.9 21 ± 14 g/m 2 22.1 ± 10.2 18.7 ± 10.6 8.0 ± 6.4 6.6 ± 6.8 6.0 ± 6.5 27.6 ± 16.0 15 ± 08 09 ± 05

±7 ± 29.6 ± 11.9 ± 7 g/m 2 ± 17.2 ± 13.2 ± 6.9 ± 4.8 ± 5.4 ± 9.2 ± 07 ± 03

Abbreviations: CK, Creatine Kinase; CK-MB, Creatine Kinase MB fraction; SPECT, single photon emission computed tomography; CMR, cardiac magnetic resonace. ⁎ Mean and SD calculated from median and Inter-quartile range.

Supplementary Table III. Outcomes Measured in the Included Studies (Markers of Myocardial Function) Study

IPoC group

Control group

LEFT VENTRICULAR EJECTION FRACTION (LVEF) 54 ± 14.5 Yang 12 Thibault 13 56 ± 8 Laskey 14 48.1 ± 4.7 Lønborg 15 53 ± 10 Xue 17 0.57 ±0.09 Sörensson 18⁎ 50 ± 9.6 Garcia 19 (Index admission) 52 ± 9 (long-term) 53 ± 8 Liu 55 ± 8 Zhao 20 (1 week) 55.19 ± 10.42 (6 months) 56.45 ± 8.54 Tarantini 24 49.4 ± 7.4 Freixa 26 (3–7 days) 42.7 ± 9.8 (6 months) 44.3 ± 10.4 52 ± 7.4 Sörensson 27⁎ Dwyer 29 (1 week) 56 ± 10 (6 months) 54 ± 08 30 Dong 55.1 ± 9.8 WALL MOTION SCORE INDEX (WMSI) Ma 10 1.2 ± 0.2 Thibault 13 1.4 ± 0.4 Xue 17 1.2 ± 0.2 Liu 1.27 ± 0.52 1.2 ± 0.17 Zhao 20 ⁎ Mean and SD calculated from median and Inter-quartile range.

44 49 45.1 53 0.47 50 43 48 47

± 16.7 ± 13 ± 4.8 ± 10 ±0.11 ± 11.1 ± 15 ± 17 ±10

51.87 ± 9.3 51.19 ± 11.22 49.9 ± 7.0 43.7 ± 8.9 47.5 ± 9.1 49 ± 10.4 52 ± 10 53 ± 07 42.9 ± 10.7 1.3 1.6 1.4 1.82 1.40

± 0.3 ± 0.4 ±0.3 ± 0.83 ± 0.32

American Heart Journal Volume 168, Number 4

Supplementary Figure 1 Funnel plot to assess publication bias for association between ischemic postconditioning in acute myocardial infarction and cardiac enzymes

Khan et al 521.e3

Supplementary Figure 2 Funnel plot to assess publication bias for association between ischemic postconditioning in acute myocardial infarction and subsequent infarct size

American Heart Journal October 2014

521.e4 Khan et al

Supplementary Figure 3 Funnel plot to assess publication bias for association between ischemic postconditioning in acute myocardial infarction and subsequent left ventricular ejection fraction

Supplementary Figure 4 Funnel plot to assess publication bias for association between ischemic postconditioning in acute myocardial infarction and subsequent wall motion score index