ST segment resolution (STR) predicts epicardial and microvascular reperfusion after primary percutaneous coronary intervention (PPCI) or thrombolysis for ST-elevation myocardial infarction. Immediate restoration of epicardial coronary flow, with improved microvascular perfusion, is much more likely with PPCI. However, the predictive value of immediate STR compared to 90 minutes after PPCI remains unknown. In 622 consecutive patients with ST-elevation myocardial infarction (mean age 59 ± 13 years), 217 had complete STR immediately after PPCI (group A), 188 had complete STR only at 90 minutes (group B), and 217 had incomplete STR at either point (group C). The primary end point was mortality and adverse cardiovascular events ([MACE] death, nonfatal repeat myocardial infarction, and heart failure). Group A had a greater left ventricular ejection fraction (53%, 47%, and 46%, p <0.001) and lower all-cause mortality (1.8%, 3.2%, and 6%, p = 0.07), lower heart failure (1.8%, 4.3%, and 7.8%, p <0.001), and MACE (5.1%, 9.6%, and 16.1%, p = 0.001) at 30 days compared to groups B and C, respectively. The rate of MACE at 1 year was 7.6%, 17.1%, and 20.2% in groups A, B, and C, respectively (p <0.001). Immediate STR independently predicted MACE (adjusted hazard ratio 0.36, 95% confidence interval 0.21 to 0.61, p = 0.001, group A vs C), and STR at 90 minutes did not. In conclusion, STR analysis performed immediately after PPCI provided superior differentiation for adverse cardiovascular events compared to STR at 90 minutes. Immediate STR should be the contemporary goal of reperfusion with PPCI.
In ST-elevation myocardial infarction (STEMI), epicardial and myocardial reperfusion is strongly predicted by the extent of ST segment resolution (STR). Incomplete STR is associated with impaired myocyte perfusion, a lack of myocardial salvage, adverse ventricular modeling, and a subsequent increase in the risk of heart failure and short- and long-term mortality. The predictive value of STR for cardiovascular outcomes after thrombolysis or primary percutaneous coronary intervention (PPCI) has been comprehensively demonstrated. STR analysis is thus recommended in the guidelines for the management of STEMI. A critical issue is the point at which the STR analysis should be performed. With lytic therapy, the precise point at which STR provides maximum prognostic differentiation of patients into high- and low-risk groups has been prospectively tested, taking into account the delay until the onset of action of the particular agent used. In the contemporary era of STEMI care with PPCI and adjunctive pharmacotherapy, prompt and sustained restoration of epicardial coronary flow with early, sustained improvement in myocyte reperfusion is much more likely than with lytic therapy. However, a limited number of studies have investigated the predictive value of STR immediately after mechanical reperfusion, and none have compared the prognosis of STR achieved at different points after PPCI. The aim of the present prospective, blinded, observational study was to assess the prognostic value of STR analysis performed immediately after PPCI compared to STR assessed at 90 minutes after PPCI for STEMI.
Methods
At Westmead Hospital (Sydney, Australia), 1,113 patients, who underwent urgent angiography from April 2004 to April 2008, were assessed and included in the present study if they had persistent ischemic chest pain of ≥20-minute duration, with ST-segment elevation of ≥1 mm in 2 contiguous leads. Patients were excluded if they had non-STEMI, unstable angina, confounding features on their electrocardiogram (ECG) such as bundle branch block or paced rhythm or a poor quality ECG, or if they had had spontaneous STR or had undergone rescue PCI (n = 329).
Patients with a history of myocardial infarction or previous revascularization with PCI or coronary artery bypass surgery (n = 115) were excluded to enable an accurate correlation among the index STEMI, STR, and measured ventricular function. Patients with missing 90-minute ECGs (n = 23) and those for whom the 90-minute ECG was not performed at the correct point were also excluded (n = 24). Thus, 622 patients with no known history of coronary artery disease, previous myocardial infarction, or previous revascularization who presented with a fresh STEMI that was treated with PPCI constituted the study population.
The operators were not aware of the present study at the PPCI. All patients received aspirin, clopidogrel, statin, β blockers, and angiotensin-converting enzyme inhibitors unless considered contraindicated by the treating physician. The use of glycoprotein inhibitors was encouraged for all patients at the intervention unless contraindicated because of the risk of bleeding. An assessment of left ventricular ejection fraction was made using radionucleotide-gated heart blood pool scanning with the patient in the left anterior oblique position for most (91%) patients or by transthoracic echocardiography for the remainder. The hospital ethics committee approved the study, and all patients gave written informed consent.
An identifier code assigned to each patient was used during electrocardiographic analysis to enable blinded assessment. The electrocardiographic data were extracted using standardized calipers in a blinded fashion by 2 investigators without knowledge of the outcomes data at separate intervals. The ST segments were assessed at 60 ms after the J point relative to the TP segment in 3 consecutive QRS-T complexes using standardized calipers to the nearest 0.1 mV in a single lead with maximum ST elevation before, immediately after (at the end of the case), and at 90 minutes after PPCI. STR was calculated by the difference between the ST-segment elevation before, immediately after, and at 90 minutes after PPCI as a percentage of the ST-segment elevation before PPCI.
Complete STR was defined as ≥70% improvement in the post versus the pre-PPCI ECGs and incomplete STR as <70% improvement. The patients were divided into 3 groups on the basis of the first point, immediate or at 90 minutes, at which complete STR was first achieved: group A, complete STR achieved immediately after PPCI; group B, complete STR achieved at 90 minutes, but not immediately, after PPCI; and group C, incomplete STR immediately or at 90 minutes after PPCI.
Two operators at the PPCI, who were unaware of the study methodology, reported the angiographic data using the Thrombolysis In Myocardial Infarction (TIMI) classification. The left ventricular ejection fraction was assessed by two nuclear medicine or imaging specialists. All patients were followed up by telephone interview or review in clinics at regular intervals of at least 1 month, 6 months, and 12 months. The in-hospital medical records were assessed for each patient. The details of the outcomes were collected by a research nurse and a co-investigator, and the outcomes data were assessed in a blinded fashion by 2 additional co-investigators who had no access to the electrocardiographic data at the assessment. The state registry was contacted to provide mortality data for all patients in the cohort and to provide details from the death certificates.
The primary end point was a composite of all-cause mortality, repeat myocardial infarction, and heart failure. Nonfatal myocardial infarction was diagnosed on the basis of a history of ischemic chest pain with new ST-segment deviation or T wave abnormalities, positive cardiac biomarker levels, and evidence of angiographic occlusion of an artery (target or nontarget vessel). Heart failure was defined as evidence of pulmonary venous and/or peripheral congestion in the medical history, examination findings, and/or imaging results, requiring ≥40 mg/day furosemide for treatment.
The Statistical Package for the Social Sciences, version 15 (SPSS, Chicago, Illinois) was used for analysis. Clinical characteristics were compared using analysis of variance, chi-square test, or Mantel-Haenszel statistics. Kaplan-Meier curves were used to compare the outcome events using the log-rank chi-square test. The baseline clinical factors listed in Table 1 were used to derive univariate predictors of the composite end point using Cox regression analysis. The univariate predictors with p ≤0.2 were entered into a multivariate model and analyzed in a backward/stepwise likelihood ratio fashion, using the Cox proportional hazards model. STR was entered as 3 independent covariates (immediate STR, 90-minute STR, and no STR at either point) and, in a separate multivariate model, as one categorical covariate with 3 co-factors (immediate, 90-minute, and no STR at either point). Statistical significance was evaluated for p <0.05.
| Variable | Group A (Immediate STR; n = 217) | Group B (90-Minute STR; n = 188) | Group C (No STR at Either Point; n = 217) | p Value |
|---|---|---|---|---|
| Age (years) | 58 ± 12 | 59 ± 12 | 60 ± 13 | 0.29 |
| Men | 79% | 76% | 80% | 0.63 |
| Hypertension | 50% | 54% | 46% | 0.26 |
| Hyperlipidemia | 57% | 53% | 51% | 0.40 |
| Diabetes mellitus | 19% | 20% | 27% | 0.10 |
| Current or past smoker | 77% | 69% | 64% | 0.08 |
| Anterior wall myocardial infarction | 21% | 44% | 68% | <0.001 |
| Systolic blood pressure ≤90 mm Hg before intervention | 9% | 6% | 4% | 0.06 |
| Multivessel coronary artery disease | 58% | 52% | 48% | 0.11 |
| Mean symptom to door time (min) | 143 ± 168 ⁎ | 136 ± 120 ⁎ | 262 ± 417 | <0.001 |
| Infarct presentation ≤3 hours | 81% | 78% | 63% | <0.001 |
| Mean door to balloon time (min) | 125 ± 64 † | 140 ± 94 † | 140 ± 67 | 0.07 |
| Within 90 min | 32% | 22% | 22% | 0.03 |
| Interval to reperfusion ‡ (min) | 268 ± 181 § | 278 ± 163 § | 398 ± 452 | <0.001 |
| Procedural time (min) | 81 ± 26 | 85 ± 31 | 84 ± 31 | 0.38 |
| Drug-eluting stent | 41% | 38% | 40% | 0.90 |
| Abciximab | 85% | 84% | 81% | 0.41 |
| Thrombolysis In Myocardial Infarction 0–2 | ||||
| Before intervention | 88% | 92% | 86% | 0.19 |
| After intervention | 0% | 2.7% | 5.5% | 0.002 |
| Intracoronary vasodilator use | 38% | 43% | 53% | 0.009 |
| Median interval to left ventricular ejection fraction assessment (days) | 3 | 3 | 3 | 0.98 |
⁎ p = 0.65, group A vs group B;
† p = 0.07, group A vs group B;
§ p = 0.57, group A vs. group B.
Results
The mean patient age was 59 ± 13 years, the mean symptom to door time was 182 ± 278 minutes, and the mean door to balloon time was 135 ± 75 minutes. Most patients underwent stenting (95.3%), with the remainder undergoing angioplasty alone (4.2%) or medical management alone (0.5%). The median follow-up duration was 397 days (mean 440 ± 265). Most patients received abciximab (83%) and 45% (n = 277) received intracoronary vasodilators (glyceryl trinitrate in 274 patients). Of the 622 patients, 217 (35%) had complete STR immediately after PPCI (group A), 188 (30%) had complete STR only at 90 minutes after PPCI (group B), and 217 (35%) had incomplete STR at either point (group C).
The 3 groups were well-matched in baseline characteristics, except that group A had a greater incidence of smokers and patients with systolic blood pressure ≤90 mm Hg on presentation ( Table 1 ). Intracoronary vasodilator use was greatest in group C. The symptom to door, door to balloon, and total reperfusion times did not significantly differ between groups A and B; however, they were significantly shorter than those for group C. Group C had a greater incidence of anterior wall myocardial infarction and TIMI flow 0 to 2 after PPCI. No significant differences were found in the procedural time; thus, the point at which immediate STR occurred was assessed among the 3 groups ( Table 1 ).
Group A had a significantly lower peak creatinine kinase and greater left ventricular ejection fraction than the other 2 groups ( Figures 1 and 2 ). At 30 days, 23 patients had died (22 cardiac deaths), 12 patients had had a nonfatal repeat myocardial infarction, and 29 patients had congestive cardiac failure. The MACE rates at 30 days were lowest in group A, followed by groups B and C ( Figure 3 ). Group A had a trend toward lower all-cause mortality and lower cardiac mortality. Group A had also significantly lower rates of heart failure; however, the rate of nonfatal myocardial infarction did not significantly differ among the 3 groups at 30 days ( Figure 3 ).
At 1 year, the MACE rates continued to be lowest in group A, followed by groups B and C, ( Figures 4 and 5 ). Overall mortality was significantly lower in group A compared to groups B and C, with a trend toward lower cardiac mortality in group A compared to groups B and C ( Figures 4 and 6 ). The incidence of heart failure was lowest in group A, followed by groups B and C ( Figure 4 ). At both 30 days and 1 year, group A had the most favorable outcomes regardless of infarct presentation of ≤3 hours, door to balloon time of ≤90 minutes, total ischemic time of ≤4 hours, and among those with TIMI flow 3 at the end of PPCI ( Table 2 ). In patients with nonanterior infarcts, the 30-day and 1-year MACE rates were lowest in group A ( Table 2 ). In patients with anterior infarcts, group A had the lowest 30-day and 1-year MACE, although this difference did not reach statistical significance until 1 year.
| Variable | Group A (Immediate STR; n = 217) | Group B (90-min STR; n = 188) | Group C (No STR at either point; n = 217) | p Value |
|---|---|---|---|---|
| 30-Day mortality and adverse cardiovascular events | ||||
| Presentation ≤3 hrs | 5.1% | 9.5% | 16.2% | 0.009 |
| Door to balloon time ≤90 min | 5.8% | 11.9% | 21.3% | 0.04 |
| Total ischemic time ≤4 hrs | 5.1% | 10.6% | 15.2% | 0.046 |
| Anterior infarction | 6.7% | 15.7% | 15.5% | 0.29 |
| Nonanterior infarction | 4.7% | 4.8% | 7.4% | 0.04 |
| Thrombolysis In Myocardial Infarction 3 flow at procedure end | 5.1% | 9.8% | 14.6% | 0.004 |
| 1-Year mortality and adverse cardiovascular events ⁎ | ||||
| Presentation ≤3 hrs | 8.3% | 20.3% | 19.5% | <0.001 |
| Door to balloon time ≤90 min | 6% | 24.4% | 25.9% | <0.001 |
| Total ischemic time ≤4 hrs | 7.1% | 20.3% | 17.7% | <0.001 |
| Anterior infarction | 9.9% | 23.2% | 12.4% | 0.001 |
| Nonanterior infarction | 7.2% | 11.8% | 22.7% | 0.001 |
| Thrombolysis in myocardial infarction 3 flow at procedure end | 7.6% | 18.5% | 17.8% | <0.001 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
