Abstract: Background. Approximately 50% of patients undergoing primary percutaneous coronary intervention (PCI) for ST-elevation myocardial infarction (STEMI) have multivessel coronary disease (MVD). Evidence on the best PCI approach for these patients is conflicting. The aim of this study is to examine Australian data from the CONCORDANCE registry to describe the practice and outcomes of patients receiving multivessel vs culprit-only PCI. Methods. Two cohorts were constructed from MVD-STEMI patients undergoing primary PCI at 41 hospitals between 2009 and 2015: culprit-only PCI (n = 587; 87%) and multivessel PCI (n = 82; 12%). Clinical characteristics were described, and the outcomes were all-cause mortality, heart failure, and myocardial reinfarction, in-hospital and at 6-month follow-up. The relative prevalence of each procedure over time was also described. Results. The patient cohorts were comparable in age, sex, and cardiovascular risk factors. Patients with higher Killip scores were more likely to receive multivessel PCI (P=.02). The multivessel group was significantly more likely to have in-hospital cardiogenic shock (P<.01), myocardial reinfarction (P=.02), cardiac arrest (P=.02), and stroke (P=.01). There was no difference in the incidence of ischemic events at 6 months, but the multivessel group had a lower rate of planned repeat revascularizations (12% vs 2%; P=.03). There was no difference in the relative frequency of multivessel vs culprit-only PCI during the observation period. Conclusions. The relative frequency of multivessel vs culprit-only PCI has not changed from 2009-2015. Index complete revascularization for STEMI-MVD patients is more likely to be performed in those with worse presentations and is associated with worse in-hospital complications.
J INVASIVE CARDIOL 2019;31(11):314-318. Epub 2019 July 15.
Key words: multivessel disease, percutaneous coronary intervention, STEMI
An estimated 50% of patients presenting with ST-elevation myocardial infarction (STEMI) and undergoing primary percutaneous coronary intervention (PCI), have multivessel coronary disease (MVD; defined as ≥50% stenosis in ≥1 non-infarct related artery).1 Patients with MVD and STEMI have a worse 30-day prognosis, usually attributed to disease extent, myocardial ischemia, or multifocal plaque instability.2 The management of this patient cohort has been a topic of debate for some time. Initially, observational data indicated no benefit in performing multivessel PCI in these patients due to safety concerns, longer procedural times, and contrast nephropathy.3,4 Recently, randomized studies have reignited the debate, where some offered data that suggest multivessel PCI in the acute setting can improve patient outcomes,5 while others have indicated that culprit-only PCI is more favorable and associated with better short-term outcomes.6 A lack of consensus about the best management of these patients remains due to the disparities between the studies’ inclusion criteria, timing of PCI, procedural techniques, and study endpoints.7
In 2011, the American College of Cardiology and American Heart Association (ACC/AHA) guidelines ranked the evidence for complete revascularization as class III evidence (harm) and only recommended the consideration of multivessel PCI in patients who are hemodynamically unstable.8 The guidelines were updated in 2015 to reflect mounting randomized data in favor of the procedure and the classification was changed to class IIb (weak evidence).7 This means that multivessel PCI may be considered in patients who are hemodynamically stable and may be performed during the index procedure or in a staged approach.7 To date, no recommendations exist in the Australian and New Zealand guidelines for the management of acute coronary syndromes (ACS).9
The objective of this study is to examine Australian data from the Cooperative National Registry of Acute Coronary Care, Guideline Adherence and Clinical Events (CONCORDANCE), a nationally representative clinical registry of patients presenting with ACS. We describe patients’ characteristics and evaluate short- and long-term outcomes in patients with acute STEMI and MVD who either receive culprit-only PCI or multivessel PCI during their index admission. An additional objective of this study is to examine the trends in the PCI strategies used for the revascularization of patients with MVD over time, considering the increasing number of published randomized trials and the changes in ACC/AHA guidelines.
Data and population. Our data come from the CONCORDANCE database, which is an ongoing electronic national longitudinal cohort registry that was established by clinicians in 2009. The detailed rationale and design of the registry have been described previously.10 The aim of the registry is to provide information on patients with ACS presenting to hospitals in geographically diverse locations in Australia.
The first 10 patients with ACS admitted to participating hospitals each month are recruited for the registry. Patients must be >18 years of age and fulfill at least one of the following criteria: electrocardiographic (ECG) changes associated with the ACS; elevated cardiac biomarkers; previous history or newly documented coronary artery disease (CAD); or two high-risk features such as diabetes or chronic kidney disease. Patients are excluded if their ACS presentation is secondary to non-cardiovascular factors such as trauma or anemia. Individual patient information is collected in a standardized electronic clinical record form (eCRF) and includes demographics, previous medical history, in-hospital management, and complications and outcomes at 6 months and at 12 or 24 months of follow-up.
Each hospital included in the CONCORDANCE registry received written approval from the appropriate Human Research Ethics Committee (HREC) to participate in data collection.10 Approval for this analysis was granted by the Sydney Local Health District (SLHD) HREC at Concord Repatriation General Hospital, New South Wales. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. In this study, the focus was on patients presenting with STEMI and MVD (defined as ≥50% stenosis in ≥1 non-infarct related artery at the time of catheterization) who received PCI as the primary treatment. Two patient groups were constructed: (1) the culprit-only PCI group; and (2) the multivessel PCI group, who received PCI for >1 vessel either during the index procedure or within the index admission.
Outcomes. The outcomes of interest were in-hospital all-cause death, heart failure, myocardial reinfarction, major adverse cardiovascular event (MACE), and any revascularization at 6-month follow-up. MACE comprised congestive heart failure, myocardial reinfarction, and cardiac death. Definitions for the variables used in this study have been described previously.10
Statistical analysis. The hospital clustering effect was adjusted for all analyses. Patient demographics, medical history, in-hospital management, and clinical events were compared between culprit-only PCI and multivessel PCI groups using the Rao-Scott Chi-square test/Fisher’s exact test for categorical variables and univariable multilevel logistic regression model for continuous variables. To assess the trends in the performance of multivessel PCI over time, a plot of multivessel procedures performed as a proportion of all primary PCI procedures in STEMI patients with MVD was constructed. P-values <.05 were considered significant. All analyses were conducted using SAS version 9.4 (SAS Institute).
Culprit-only PCI vs multivessel PCI in the index admission. A total of 669 patients were included in the analysis. Of those, eighty-two patients (12%) received multivessel PCI, while the rest received culprit-only PCI. Table 1 outlines the baseline characteristics of the patients. Most of the study population (79%) was male, and patients were comparable in terms of age and background risk factors, such as diabetes (21% vs 23%; P=.58), dyslipidemia (45% vs 40%; P=.41), and smoking status (37% vs 28%; P=.08). In addition, clinically modest non-significant differences were detected in their previous history of cardiovascular disease, including previous MI, previous PCI, and congestive heart failure. One statistically significant difference was noted between the Killip scores, with the multivessel PCI group demonstrating higher Killip scores at presentation vs the culprit-only PCI group (P=.02).
The in-hospital complications revealed several clinical differences between the culprit-only PCI and multivessel PCI groups that reached statistical significance (Table 2). Specifically, the differences were found in cardiogenic shock (5% vs 15%, respectively; P<.01), myocardial reinfarction (3% vs 9%, respectively; P=.02), cardiac arrest (5% vs 12%, respectively; P=.02), and stroke (1% vs 5%, respectively; P=.01). The pooled MACE variable was also significantly greater in the multivessel group (17% in the culprit-only PCI group vs 27% in the multivessel PCI group; P=.04). However, in-hospital mortality did not differ between the two groups (4% in the culprit-only PCI group vs 7% in the multivessel PCI group; P=.25).
Outcomes at 6 months. Follow-up data were available for 76% of the culprit-only PCI group and 67% of the multivessel PCI group. No differences were found in cardiac death rate (1% vs 0%, respectively; P=.50), myocardial reinfarction (2% vs 4%, respectively; P=.63), congestive heart failure (3% vs 5%, respectively; P=.39), or MACE (6% vs 9%, respectively; P=.36).
The need for repeat revascularization between discharge and 6-month follow-up is presented separately in Table 4. It shows that the culprit-only PCI group was numerically more likely to receive any revascularization by 6-month follow-up compared with the multivessel PCI group, although the difference did not reach statistical significance (14% vs 7%, respectively; P=.08). When looking at the rate of planned procedures alone, the culprit-only PCI group had significantly higher rates of planned revascularization at 6-month follow-up vs the multivessel PCI group (12% vs 2%, respectively; P=.03), perhaps corresponding to a staged approach.
Trends in MVD revascularization using PCI over time. Figure 1 shows the proportion of multivessel PCI procedures performed on STEMI patients with MVD and undergoing primary PCI, as recorded in the CONCORDANCE registry since 2009. Although there was a numerical trend in increasing frequency of multivessel relative to culprit-only procedures, no significant changes were noted over 6 years as a proportion of all MVD-STEMI primary PCIs.
In this study, we offer insight into the practice of PCI in patients with STEMI and MVD in Australia. Our registry data suggest that patients presenting with higher Killip scores are more likely to receive multivessel PCI than a culprit-only procedure. In the index admission, patients who received multivessel PCI displayed worse in-hospital complications, but did not have significantly greater cardiac morbidity or mortality at 6-month follow-up.
Short-term observations in this cohort are not surprising. Prior to the availability of any randomized studies, multivessel PCI during an acute STEMI presentation was attempted more frequently in patients who presented with hemodynamic compromise or severe symptoms, as a “last resort” to achieve reperfusion and prevent death. Our data appear to reflect this, as the multivessel PCI patients presented with higher Killip scores and went on to have complications relating to their hemodynamic state including cardiogenic shock and cardiac arrest (Table 2).
In patients with shock, we know that coronary reperfusion, whether through a single or multivessel procedure, is a major predictor of survival.11 We found a numerical increase in hospital deaths in the multivessel PCI group, perhaps reflecting the increased proportion of patients with shock. Consistent with this, the recently published results of the CULPRIT-SHOCK trial showed that STEMI patients with MVD and shock who received index multivessel PCI had significantly higher death rates at 30 days than those undergoing culprit-lesion PCI only.6 However, we acknowledge that the absolute number of shocked patients in our study was very small and the data should be interpreted with that in mind.
Despite differences in short-term complications, our 6-month follow-up data show no significant differences between the two groups in mortality, heart failure, cardiac death, or MACE (Table 3). Previously, studies have shown that MACE rates were significantly lower in the multivessel group at long-term follow-up.5,12,13 When comparing our analysis with that of Gershlick et al, we noted that they included repeat revascularization (using either PCI or coronary artery bypass grafting) as a component of their MACE variable rather than analyzing it separately.5 Apart from that, the other MACE components were identical to our own. Wald et al also found that MACE rate was significantly lower in the multivessel PCI group compared with the culprit-only PCI group.12 However, their study excluded patients who were in cardiogenic shock as well as those with chronic total occlusions or a major left main artery stenosis, citing the need for coronary artery bypass grafting in these patients.12 Comparable exclusion criteria were used in the DANAMI-3 PRIMULTI study by Engstrøm et al, and their results also favored multivessel PCI; however, the patients in the multivessel group received PCI in a staged approach, with fractional flow reserve guidance 2 days after the index procedure.13 It is possible that the variations in methodology contributed to our differing findings.
The only significant difference we observed was in the rate of planned revascularization procedures between discharge and 6 months, with the culprit-only group more likely to receive a planned procedure (P=.03). Again, this result is somewhat expected, as patients who receive culprit-only PCI would presumably have other lesions treated in a staged approach, based on their initial diagnostic angiography. A limitation to our data is that patients who had staged PCI during their index admission could not be separated from the cohort in our registry and therefore could not be studied separately.
A recent meta-analysis by Tarantini et al looking at the timing of PCI examined data from 32 cohort studies involving more than 54,000 patients. The study compared three different protocols of MVD management following the acute STEMI presentation: (1) culprit-only PCI; (2) multivessel PCI during the index procedure; and (3) multivessel PCI in a staged approach during the first 30 days from presentation. The results were in favor of staged multivessel PCI, as it was associated with lower short-term (30-day) and long-term (6-month) mortality compared with the other two protocols.3 However, the randomized trials discussed above suggested that both index and staged multivessel PCI may have favorable outcomes by reducing cardiac-related mortality (at 23-month follow-up),12 risk of reinfarction (at 12-month follow-up),5 and the need for repeat PCI (at a median of 27 months of follow-up).13 The best PCI approach for patients with MVD remains unresolved. Data from these described randomized trials became available during the latter part of our data collection period. It appears as though these trials had limited impact on Australian practice, as our analysis of the trends in Australian practice of primary PCI revascularization in STEMI patients with MVD showed no significant change over the 6-year period.
Study limitations. Data from the CONCORDANCE registry records vessel stenoses as ≥50% using visual estimation by the operator and not through a standardized method at all participating centers. Also, chronic total occlusions are not separately recorded and may confound the data because of their added complexity. Six-month follow-up was incomplete and it is possible that events occurred in patients who could not be contacted. We expect the bias here would affect both groups equally. Our data are observational and therefore are susceptible to reporting bias.
Our study showed that immediate complete revascularization for STEMI patients with MVD is more likely to occur in patients with hemodynamic compromise at presentation (higher Killip class); these patients have worse short-term outcomes, but no difference in mortality or cardiac morbidity compared with culprit-only PCI patients at 6 months. Conversely, culprit-only PCI is associated with a higher likelihood of planned repeat revascularization at 6 months. No changes were observed in the Australian multivessel PCI practices over the study time. These findings reflect the ongoing debate in the literature and the lack of robust randomized evidence to trigger changes in clinical practice or updates to the guidelines and recommendations.
1. Park D, Clare RM, Schulte PJ, et al. Extent, location, and clinical significance of non–infarct-related coronary artery disease among patients with ST-elevation myocardial infarction. JAMA. 2014;312:2019-2027.
2. Bates ER, Tamis-Holland JE, Bittl JA, O’Gara PT, Levine GN. PCI strategies in patients with ST-segment elevation myocardial infarction and multivessel coronary artery disease. J Am Coll Cardiol. 2016;68:1066-1081.
3. Tarantini G, D’Amico G, Brener SJ, et al. Survival after varying revascularization strategies in patients with ST-segment elevation myocardial infarction and multivessel coronary artery disease: a pairwise and network meta-analysis. JACC Cardiovasc Interv. 2016;9:1765-1776.
4. Di Pasquale G, Filippini E, Pavesi PC, Tortorici G, Casella G, Sangiorgio P. Complete versus culprit-only revascularization in ST-elevation myocardial infarction and multivessel disease. Intern Emerg Med. 2016;11:499-506.
5. Gershlick AH, Khan JN, Kelly DJ, et al. Randomized trial of complete versus lesion-only revascularization in patients undergoing primary percutaneous coronary intervention for stemi and multivessel disease: the CvLPRIT trial. J Am Coll Cardiol. 2015;65:963-972.
6. Thiele H, Akin I, Sandri M, et al. PCI strategies in patients with acute myocardial infarction and cardiogenic shock. N Engl J Med. 2017;377:2419-2432.
7. Levine GN, Bates ER, Blankenship JC, et al. 2015 ACC/AHA/SCAI focused update on primary percutaneous coronary intervention for patients with ST-elevation myocardial infarction: an update of the 2011 ACCF/AHA/SCAI guideline for percutaneous coronary intervention and the 2013 ACCF/AHA guideline for the management of ST-elevation myocardial infarction. J Am Coll Cardiol. 2016;67:1235-1250.
8. Levine GN, Bates ER, Blankenship JC, et al. 2011 ACCF/AHA/SCAI guideline for percutaneous coronary intervention: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines and the Society for Cardiovascular Angiography and Interventions. Circulation. 2011;124:e574-e651.
9. Chew DP, Scott IA, Cullen L, et al. National Heart Foundation of Australia & Cardiac Society of Australia and New Zealand: Australian clinical guidelines for the management of acute coronary syndromes 2016. Heart Lung Circ. 2016;25:895-951.
10. Aliprandi-Costa B, Ranasinghe I, Turnbull F, et al. The design and rationale of the Australian Cooperative National Registry of Acute Coronary Care, Guideline Adherence and Clinical Events (CONCORDANCE). Heart Lung Circ. 2013;22:533-541.
11. Webb JG, Lowe AM, Sanborn TA, et al. Percutaneous coronary intervention for cardiogenic shock in the SHOCK trial. J Am Coll Cardiol. 2003;42:1380-1386.
12. Wald DS, Morris JK, Wald NJ, et al. Randomized trial of preventive angioplasty in myocardial infarction. N Engl J Med. 2013;369:1115-1123.
13. Engstrøm T, Kelbæk H, Helqvist S, et al. Complete revascularisation versus treatment of the culprit lesion only in patients with ST-segment elevation myocardial infarction and multivessel disease (DANAMI-3—PRIMULTI): an open-label, randomised controlled trial. Lancet. 2015;386:665-671.
From the 1University of Sydney, New South Wales, Australia; 2the George Institute for Global Health, New South Wales, Australia; 3Sydney Local Health District, Sydney, Australia; 4Flinders University, Adelaide, Australia; 5Royal Prince Alfred Hospital, New South Wales, Australia; 6Liverpool Hospital, New South Wales, Australia; 7Concord Repatriation General Hospital, New South Wales, Australia; and 8Royal Brisbane and Women’s Hospital. Queensland, Australia.
Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Brieger reports travel grant/speaker fees from Pfizer/BMS; travel grant from Boehringer Ingelheim; researh support from Astra Zeneca. Dr Chew reports research support from Roche Diagnostics and Edwards Lifesciences; speaker fees from Astra Zeneca. The remaining authors report no conflicts of interest regarding the content herein.
Manuscript submitted February 25, 2019 and accepted March 4, 2019.
Address for correspondence: David Brieger, MBBS, PhD, MMed (ClinEpi), 7 Concord Repatriation General Hospital, The University of Sydney, NSW, Australia. Email: David.Brieger@health.nsw.gov.au