Article Text
Abstract
Objective ST-elevation myocardial infarction (STEMI) is frequently associated with reciprocal ST-segment depression in contralateral ECG leads. However, the relationship of the resolution of ST-segment depression (STD-R) with myocardial damage is unknown and the potential prognostic value incompletely understood. We sought to evaluate the association between STD-R and markers of myocardial injury as well as to determine the prognostic impact of STD-R in patients with acute reperfused STEMI.
Methods We enrolled 611 patients with STEMI in this multicentre cardiac magnetic resonance (CMR) study. STD-R, defined as either worsened (<0%), incomplete (0–50%) or complete (≥50%), was determined 90 min after primary percutaneous coronary intervention (PCI). Patients underwent CMR in median 3 (2–4) days after infarction. Major adverse cardiac events (MACE) were defined as a composite of death, reinfarction and new congestive heart failure within 12 months after enrolment.
Results Patients with worsened or incomplete STD-R (n=148 (24.2%)) had a significantly larger area at risk (42 (31–50) vs 37 (29–52) vs 34 (24–46) %LV, p=0.001), larger infarct size (20 (13–30) vs 17(10–26) vs 16 (8–24) %LV, p=0.003), larger extent of microvascular obstruction (0.6(0–3.4) vs 0.4 (0–2.4) vs 0.0 (0–1.4) %LV, p=0.003), and a lower LVEF (46 (39–54) vs 48 (40–56) vs 52 (45–58) %, p<0.001). MACE rate (n=37 (6%)) was significantly higher in patients with worsened (n=10 (19%)) or incomplete STD-R (n=7 (7%)) than in patients with complete STD-R (n=20 (4%), p<0.001). In multivariate Cox regression analysis, categorised STD-R emerged as an independent predictor of MACE at 12 months after adjusting for clinical variables (p=0.007).
Conclusions Patients with acute STEMI and worsened or incomplete STD-R after PCI show a more pronounced myocardial as well as microvascular damage as detected by CMR with subsequent independent prognostic information on MACE over a 12-month follow-up period.
Statistics from Altmetric.com
Introduction
Numerous studies have shown that the extent of early ST-segment elevation resolution (STR) after primary percutaneous coronary intervention (PCI) evaluated on standard 12-lead ECG is an indicator of successful reperfusion and future cardiovascular events in patients with ST-elevation myocardial infarction (STEMI).1 Although early studies on patients with STEMI already reported that concomitant ST-segment depression is common,2 ,3 only few assessed the prognostic value of ST-segment depression resolution (STD-R) after reperfusion therapy.4 ,5 Recently, a subanalysis of the Assessment of Pexelizumab in Acute Myocardial Infarction (APEX-AMI) trial showed that STD-R after PCI provides strong independent prognostic information that is incremental to STR.6 In line with these data, Wong et al7 found that in patients with inferior STEMI treated with fibrinolysis STD-R is prognostically more important than standard STR. However, the underlying mechanism for the strong association between STD-R and adverse outcome remains poorly understood and most prognostic data are only available for relatively short follow-up periods.
Contrast-enhanced cardiac magnetic resonance (CMR) imaging is the most accurate method for the visualisation of myocardial injury in vivo.8 This modality is therefore uniquely suited to elucidate the potential association between STD-R and myocardial damage as well as reperfusion injury after PCI for acute STEMI.
In the present study, we therefore aimed to investigate the relationship of STD-R early after PCI with myocardial and microvascular damage as measured by CMR and to assess the prognostic value of incomplete STD-R in a long follow-up period (12 months) after infarction in a large multicentre STEMI cohort.
Methods
Study population
This multicentre CMR study was a substudy of the Abciximab Intracoronary versus intravenous Drug Application in ST-Elevation Myocardial Infarction (AIDA STEMI) trial. Protocol and principal results of the AIDA STEMI trial have been published previously.9–11 The protocol of the predefined CMR substudy of AIDA STEMI has been reported in detail previously.12 This substudy recruited 795 consecutive patients with STEMI at eight sites. For the purpose of the current analysis, only patients receiving a 12-lead ECG at baseline as well as 90 min following PCI were eligible. Out of 751 patients with STEMI with an analysable CMR and ECG, 140 (n=18.6%) displayed no ST-segment depression on admission and were therefore excluded from the current analysis. Consequently, data of 611 patients with STEMI were available for final analysis (figure 1).
ECG analysis
According to the prespecified study protocol, standard 12-lead ECGs (voltage 10 mm/mV, speed 25 mm/s or 50 mm/s) were recorded in each case prior to and 90 min after PCI. Two experienced cardiologists blinded to clinical and CMR data analysed all 12 leads for ST-segment depression at both time points at the ECG core laboratory (University of Leipzig—Heart Center, Leipzig, Germany). The amount of ST-segment depression was measured manually with calliper 20 ms after J point to the nearest 0.05 mV. A consensus decision was made in case of divergent results between the two readers. The STD-R in percentage was calculated according to the following formula:
We identified several patients with more or less negative STD-R values with worsened ST depression after reperfusion. Grouping STD-R into three categories is therefore reasonable. In addition to the two categories based on previous literature6—(1) incomplete (non-negative) STD-R defined as <50% and (2) complete STD-R (50–100%)—we defined category (3) as worsened STD (<0%).
CMRI
CMR scans were performed on a 1.5 Tesla or 3.0 Tesla scanner using standardised sequences as previously described.9 Evaluation of images was performed using standard CMR postprocessing software (cmr42, Circle Cardiovascular Imaging, Calgary, Alberta, Canada). Readers blinded to study data performed the image analysis at the CMR core laboratory (University of Leipzig—Heart Center). Core laboratory analysis guarantees excellent reproducibility and low interobserver as well as intraobserver variability for the measurement of infarct size (IS) and myocardial salvage (MS).13 The measurements of the area at risk (AAR), IS, MS and microvascular obstruction (MVO) were expressed as the percentage of LV volume (% LV).12 Myocardial salvage index (MSI) and intramyocardial haemorrhage were determined as recently described.12 ,14
Clinical follow-up
The primary clinical endpoint of this substudy was the incidence of major adverse cardiac events (MACE), defined as a composite of all-cause death, reinfarction and new congestive heart failure at 12 months after enrolment. Detailed endpoint definitions are reported elsewhere.9 ,10
Statistical analysis
Continuous variables are presented as mean (SD) or median with IQR and categorical variables as number and percentage. In accordance, analysis of variances and Jonckheere–Terpstra test for ordered alternative were used to determine differences in continuous variables between groups. Counts were compared between groups by Armitage’s trend for proportions. We applied exact tests where necessary. Event-free survival was depicted by means of Kaplan–Meier method and tested by log-rank test. We performed multiple Cox regression for two reasons: first, to adjust for confounders; and second, to look for potential predictors of the composed endpoint (MACE). For this aim, we used the LASSO algorithm (‘least absolute shrinkage and selection operator’) in the R package penalised.15 Having 37 MACE events in follow-up, a Cox regression model with up to three covariates is appropriate in accordance with the rule of thumb ‘one variable for 10 events’. The very flexible LASSO method enables the selection of a prespecified number of variables without considering p values and gives relatively stable models. By calculating the profile function, we found an optimal lambda parameter. Then a Cox (linear) regression model was fitted by penalised regression. We reduced the number of non-zero coefficients by increasing lambda until the appropriate number of variables left. Finally, we built a prognosis model with these parameters. Univariate and multivariate linear regression was performed to assess whether STD-R is an independent predictor of IS and MVO assessed by CMR. Multivariate regression was performed using only variables with a p value <0.05 in univariate regression analyses. For univariate analyses, all variables listed in table 1 and STD-R were investigated. The incremental additive information of STD-R over standard STR for the prediction of MACE was assessed with c-statistics.16 A two-sided α=5% was determined overall. Statistical analysis was performed with SPSS V.22.0 (IBM, Armonk, New York, USA) and MedCalc V.12.2.1 (Ostend, Belgium).
Results
Patient characteristics
Post-reperfusion ECGs were obtained at median time 111 min (IQR 80–169) after PCI. The absolute number of leads with ST-elevation in the ECG before and after PCI were 5 IQR (3–6) and 3 (IQR 2–5), respectively. STD was detected in median in 3 (IQR 2–5) leads before PCI and in 1 (IQR 0–3) lead after PCI. The median STR-D was 88% (IQR 50–100). Worsened STD-R was present in 52 (8.5%) patients, whereas STD-R of <50% occurred in 148 (24.2%) patients. Baseline characteristics of the entire study population as well as their association with STD-R are summarised in table 1.
Patients with lower STD-R trended to be older (p=0.07) and had a more unfavourable cardiovascular risk profile (hypertension (p=0.09) and diabetes mellitus (p=0.01)). Admission heart rate (p=0.04), Killip class at admission (p=0.001) and postprocedural Thrombolysis In Myocardial Infarction (TIMI) flow grade (p<0.001) were also worse in patients with lower STD-R. In addition, there was a higher use of thrombus aspiration and intra-aortic balloon pump in both groups of STD-R<50% (p=0.002 and p<0.001, respectively).
STD-R and CMR parameters
Patients underwent CMRI within a median of 3 (IQR 2–4) days after STEMI. There was no difference regarding the time point of CMRI between the three STD-R groups (median 3 days, IQR 2–4 days in all groups, p=0.82). Similarly, there was no difference in time from PCI to post-PCI ECG between both STD-R groups (p=0.39). CMR findings and their relation with STD-R are depicted in table 2.
Lower STD-R was significantly associated with a larger AAR (p=0.001), larger IS (p=0.003), higher LV end-systolic volume (p=0.001), lower LVEF (p<0.001) and a larger extent of MVO in % LV (p=0.003). In addition, STD<0% was significantly associated with lower MSI (p=0.015). STD-R was an independent predictor of MVO (table 3), but not of IS in multivariable regression analysis.
STD-R and clinical outcome
At 12 months, there were 37 (6%) incidences of MACE, including 17 deaths (3%), 10 reinfarctions (2%) and 10 new congestive heart failure events (2%). Patients with negative STD-R had 10 events (19%), patients with STD-R≤50% had 7 (7%) and those with 50–100% STD-R 20 events (only 4%), demonstrating a graded relationship across the STD groups with MACE-free survival (log-rank test p<0.001) (figure 2). Multivariate Cox regression analysis, omitting CMR parameters, revealed age (p=0.001), heart rate (p=0.018) and categorised STD-R (p=0.007) as independent predictors of MACE (table 4A). When CMR parameters (IS) where introduced in addition to age as well as categorised STD-R, STD-R remained an independent predictor for MACE (table 4B).
Lower STD-R was also significantly associated with higher mortality (4 (8%), 3 (3%) vs 10 deaths (2%), p=0.03) and a higher incidence of readmission for acute heart failure (5 (10%), 5 (5%) vs 8 (2%), p=0.001). The occurrence of reinfarction was not significantly different between the groups (3 (6%), 1 (1%) vs 8 (2%), p=0.16). The risk for an event was especially high for patients with worsened STD (death: OR 3.74, 95% CI (1.09 to 11.0), hospitalisation for heart failure: OR 4.44 (1.52 to 13.0)). Receiver operating characteristic curve analysis further illustrated that STD-R is a strong indicator of adverse clinical events. The inclusion of STD-R in addition to STR resulted in a significant increase of the c-statistics from 0.61 (95% CI 0.57 to 0.65) to 0.72 (95% CI 0.68 to 0.75, p=0.02), thus demonstrating an additive prognostic value of STD-R over STR.
Discussion
This is the first CMR multicentre study investigating the association between STD-R and IS as well as reperfusion injury and its impact on hard clinical events among patients undergoing PCI for acute STEMI. The main findings can be summarised as follows: (1) patients with worsened or incomplete STD-R after PCI had larger infarcts, but also, and possibly of greater importance, more MVO and less MS compared with patients with complete STD-R. (2) STD-R was a strong and independent predictor of MACE at 12 months following STEMI when adjusting for clinical risk factors. (3) STD-R has an incremental prognostic value in addition to the widely used ECG parameter STR. Consequently, our findings demonstrate a pathophysiological link between incomplete STD-R and increased myocardial injury with subsequent unfavourable outcome in patients with STEMI.
Incidence and determinants of STD-R in STEMI
Until now, only a few studies have addressed STD-R as a distinct ECG metric of reperfusion success in patients with acute reperfused STEMI. Moreover, most of these studies were performed in the pre-PCI era and they were hampered by small sample sizes and single-centre observations. Early fibrinolytic reperfusion studies revealed an incidence of incomplete STD-R of approximately 15%.5 According to the largest study to date using contemporary immediate PCI therapy as reperfusion strategy, one out of four patients with STEMI develops poor STD-R (defined as STD-R≤50%). Our results confirm that incomplete STD-R is a common finding with an occurrence of approximately 25% in patients with STEMI reperfused by PCI. However, the underlying mechanism of concomitant ST-segment depression and particularly its resolution after PCI in patients with STEMI is highly controversial. It has been speculated that reciprocal ST-segment depression may reflect a more extensive infarction,4 ,17 ,18 additional ischaemia beyond the infarct zone,18–21 more diffuse coronary artery disease6 ,22 or simply a benign mirror projection of the ST-segment elevation without significant clinical relevance.23 ,24
Recently, Kidambi et al reported in a small single-centre CMR study (n=35) that patients with STEMI with reciprocal changes on admission ECG have a larger mass of ischaemic myocardium at risk and higher extent of salvaged myocardium as visualised by CMR.25 In contrast, there was no correlation with IS or MVO and no evidence for remote ischaemia. Unfortunately, Kidambi et al did not evaluate the association of the recovery of ST-segment depression with myocardial damage after PCI. Our study is therefore the first investigation that comprehensively evaluates the correlation between STD-R and CMR markers of myocardial and reperfusion injury. We observed that incomplete STD-R has a complementary predictive value on myocardial function, size and extent of infarction, as well as microvascular damage. In addition, multivariate regression analysis showed that STD-R is an independent predictor of MVO. Recent data underscore that the integrity and maintenance of microvascular perfusion during acute infarction is of particular prognostic importance.26 ,27 Consequently, we suggest that early STD-R after PCI not only signifies effective epicardial blood flow, but more importantly also adequate microvascular perfusion. Incomplete STD-R may therefore better express reperfusion injury at the myocardial tissue level than STR or the simple presence of concomitant ST-depression.
Prognostic value of STD-R
Concomitant ST-segment depression and incomplete STD-R has been associated with poor clinical outcomes.5–7 Moreover, the APEX-AMI trial demonstrated for the first time that STD-R≤50% provided prognostic information that is incremental to STR, the most widely used surrogate ECG marker of successful tissue reperfusion in STEMI. We recognised that a significant portion of patients had even worsened STD-R after reperfusion, and consequently categorised STD-R in three categories. In line with previous studies, we could confirm that worsened and/or STD-R≤50% was an independent predictor of hard clinical events with incremental prognostic value beyond clinical risk factors and STR. Additionally, the follow-up time of 12 months in the present study is longer compared with most of the previously reported trials,5 ,6 highlighting the mid-term prognostic value of STD-R. Our study complements previous findings by demonstrating the pathophysiological link between incomplete STD-R with more severe myocardial damage and consequently worse clinical outcome.
Since CMRI is not applicable to all patients with STEMI in daily clinical routine determination of STD-R might be a more feasible alternative to immediately estimate the extent of successful myocardial and microvascular reperfusion after PCI. Furthermore, this simple tool has the potential to immediately identify patients who might benefit from more intensified treatment strategies. Whether these subsets of high-risk patients benefit from additional therapeutic interventions needs to be determined in future prospective studies. Moreover, the current results emphasise the use of STD-R as a simple and strong surrogate endpoint in clinical trials evaluating novel reperfusion strategies.
Limitations
Our study has some limitations. In contrast to other trials where ECG was performed 30 min or 3 h post-PCI, we evaluated STD-R at 90 min after PCI therapy. Although STD-R was a strong predictor of hard clinical events in all studies, the optimal time point for the assessment of STD-R after PCI remains to be established. Recent data from STR studies have shown that patients with a very early resolution of ST-segment elevation after PCI (<30 min) have a better myocardial function and smaller infarcts compared with patients with late or no STR (30 min or later).28 Therefore, to elucidate potential differences, studies focusing on the direct comparison between different time points of STD-R assessment are warranted. Based on the inclusion and exclusion criteria of this multicentre study and treatment in specialised high-volume hospitals, our results may not be not generalisable to other STEMI populations, for example, late STEMI presenters (>12 h).
In conclusion, worsened or incomplete STD-R after PCI for STEMI is correlated with markers of severe myocardial damage with subsequently increased risk of MACE 12 months after infarction. STD-R after PCI provides incremental prognostic information to clinical risk factors and STR. Consequently, STD-R is a non-invasive parameter of successful reperfusion that mirrors tissue and microvascular perfusion, which might improve early risk stratification in patients with STEMI.
Key messages
What is already known on this subject?
Concomitant ST-segment depression is common in patients presenting with acute ST-elevation myocardial infarction (STEMI). In addition, patients with incomplete resolution of ST-segment depression (STD-R) after primary percutaneous coronary intervention (PCI) face an increased risk for future cardiovascular events. However, the underlying mechanism between STD-R and adverse outcome is incompletely understood.
What might this study add?
This study demonstrates that worsened or incomplete STD-R after PCI is significantly related with a larger area at risk, larger infarct size, less myocardial salvage and larger amount of microvascular obstruction as measured by cardiac MRI. Our results highlight the pathophysiological association of incomplete STD-R with more severe damage of the myocardium and microvasculature with subsequent adverse prognosis in these patients.
How might this impact on clinical practice?
STD-R is a non-invasive parameter of successful reperfusion that mirrors tissue and microvascular perfusion, which might improve early risk stratification in patients with STEMI. Moreover, the results emphasise the use of STD-R as a simple and strong surrogate endpoint in clinical trials evaluating novel reperfusion strategies.
References
Footnotes
Contributors SJR: conceived the study; analysis and interpretation of data; drafting of the manuscript. AB: interpretation of the data; drafting of the manuscript and revising critically for important intellectual content. KPR: analysis and interpretation of the data; revising critically for important intellectual content. CE: analysis and interpretation of the data; revising critically for important intellectual content. SD: analysis and interpretation of the data; revising critically for important intellectual content. MM: statistical analysis and interpretation of the data; revising critically for important intellectual content. BM: interpretation of the data; revising critically for important intellectual content. JP: analysis of the data; revising critically for important intellectual content. HT: interpretation of the data; revising critically for important intellectual content. IE (Corresponding author): conceived the study; analysis and interpretation of data; drafting of the manuscript; revising critically for important intellectual content.
Competing interests None declared.
Patient consent Obtained.
Ethics approval Ethics committee.
Provenance and peer review Not commissioned; externally peer reviewed.