Although a fibrinolytic pharmacoinvasive strategy is recommended for patients with ST elevation myocardial infarction (STEMI) unable to undergo timely primary percutaneous coronary intervention (PCI), there are limited data addressing outcomes specific to those with successful or unsuccessful pharmacologic reperfusions. Accordingly, we evaluated a contemporary pharmacoinvasive strategy for failed and successful reperfusions within the STrategic Reperfusion Early After Myocardial infarction study. Of 1,823 per-protocol–treated patients with STEMI, we examined clinical outcomes and angiographic and electrocardiographic metrics in 3 groups as follows: fibrinolysis requiring rescue (rescue, n = 348), fibrinolysis with scheduled angiography (scheduled, n = 516), and primary PCI (n = 927). Compared with pharmacoinvasive patients undergoing scheduled angiography, rescue patients were more likely to have anterior wall myocardial infarction, more baseline ST-segment elevation and deviation, as well as Q waves in the distribution of their ST elevation. Residual ST elevation ≥2 mm 30 minutes after angiography occurred in 27.9%, 7.9%, and 24.8% of patients who underwent rescue, scheduled, and primary PCI, respectively. Thirty-day composite event rates (all-cause death, cardiogenic shock, heart failure, and reinfarction) were 18.7%, 5.5%, and 13.9%; all-cause death: 6.1%, 2.1%, and 3.9%; cardiogenic shock: 7.5%, 2.0%, and 5.4%; heart failure: 11.8%, 2.3%, and 7.6%; and reinfarction: 1.5%, 1.4%, and 2.2%, for rescue, scheduled, and primary PCI, respectively. Compared with successfully reperfused patients undergoing scheduled angiography, the adjusted relative risk of the primary outcome was 2.92 (95% confidence interval 1.92 to 4.45) in rescue patients. In conclusion, pharmacoinvasive-treated patients requiring rescue angiography had greater baseline risk with more co-morbidities and worse 30-day outcomes compared with successful fibrinolytic-treated patients. Residual ST elevation after reperfusion assists in defining prognosis.
We have previously demonstrated, in patients with ST elevation myocardial infarction (STEMI) presenting within 3 hours of symptom onset (who could not undergo primary percutaneous coronary intervention [PCI] within 60 minutes), that a pharmacoinvasive strategy that includes tenecteplase (TNK), urgent PCI for failed reperfusion, and angiography within 24 hours in the remaining fibrinolytic-treated patients achieved a comparable 30-day composite end point of all-cause death, cardiogenic shock, congestive heart failure (CHF), and reinfarction compared with timely contemporary primary PCI. There was a slight excess of intracranial bleeding with pharmacoinvasive therapy, but the rates of nonintracranial bleeding were similar. To enhance the limited existing knowledge in patients managed with a contemporary pharmacoinvasive strategy, in the present report, we examine the characteristics and outcomes of pharmacoinvasive-treated patients with STEMI managed by protocol-defined angiography after unsuccessful fibrinolysis and compare them with those experiencing successful reperfusion and subsequently undergoing scheduled angiography. Additionally, we explore what insights are provided by sequential core laboratory electrocardiographic (ECG) data relating to reperfusion success, the outcomes of the Strategic Reperfusion Early After Myocardial infarction (STREAM) study, and their implications for clinical practice.
Methods
The STREAM study ( www.clinicaltrials.gov [ NCT00623623 ]) protocol and primary result have been previously reported. Briefly, patients presenting within 3 hours of symptom onset were randomized to either fibrinolysis with a protocol-defined pharmacoinvasive strategy or primary PCI. Specifically, bolus TNK, aspirin, clopidogrel, and enoxaparin were administered according to guideline recommendations and followed by either rescue coronary intervention (hereafter referred to as rescue) or scheduled coronary angiography within 6 to 24 hours (hereafter referred to as scheduled angiography). Primary PCI was conducted with expedited transfer from the point of randomization and early initiation of aspirin, clopidogrel (300 or 600 mg), and antithrombotic therapy based on best standard practice. This report is based on a per-protocol treatment analysis according to the treatment that was received ( Figure 1 ). Patients who did not have confirmed STEMI, did not receive fibrinolysis, or did not undergo cardiac catheterization were excluded. Before angiography, there were 12 deaths in the scheduled angiography group, 5 in the primary PCI group, and none in those who underwent rescue PCI.
The site investigators defined the infarct-related culprit artery, baseline Thrombolysis In Myocardial Infarction (TIMI) flow grade, and postprocedure TIMI flow grade using standard definitions. The postprocedure TIMI flow grade is presented after coronary angiography and separately in those who received PCI (post-PCI). Multivessel disease was visually identified by site investigators in patients with >1 coronary artery stenosis >50% or with >50% stenosis in the left main coronary artery.
The need for rescue was identified by local investigators according to whether there was <50% ST-elevation resolution in the ECG lead with maximal ST elevation 90 minutes after TNK, hemodynamic instability, or refractory ventricular arrhythmias as mandated by study protocol. An additional subset of patients were identified by clinical and ECG evidence as initial successful reperfusion but then exhibited recurrent ECG evidence of ST elevation or clinical indications within 6 hours of TNK; these were deemed “urgent angiography” for purposes of the present report.
Electrocardiograms were collected at baseline, 90 minutes after TNK, and 30 minutes after catheterization (including after PCI in those undergoing PCI) in the pharmacoinvasive arm and 30 minutes after catheterization or PCI (if performed) in the primary PCI arm. They were interpreted by experienced ECG readers at the Canadian VIGOUR Centre ECG Core Laboratory (Edmonton, Alberta, Canada). The prespecified ECG metrics included a baseline Q wave in the infarct distribution; sum ST-segment deviation at baseline and after TNK, angiography, and PCI; worst lead ST elevation at baseline and after TNK, angiography, and PCI; worst lead residual ST elevation after TNK, angiography, and PCI, and these were acquired according to methods previously described. Worst lead residual ST elevation has been previously demonstrated to have excellent prognostic utility in patients with STEMI undergoing primary PCI, and assessment of 50% ST-elevation resolution is recommended as an assessment metric for fibrinolytic patients in STEMI guidelines. Left bundle branch block, ventricular rhythm, paced rhythm, and incomplete or poor quality electrocardiograms were excluded from analysis of ECG metrics.
Baseline patient demographics, key timelines, and ECG and procedural characteristics are presented as percentages for categorical variables, and differences among groups were tested by the chi-square test; Fisher’s exact test was applied when the cell count was <5 patients. Continuous variables are summarized as medians with twenty-fifth and seventy-fifth percentiles, and differences among groups were tested through the Wilcoxon rank sum test.
Factors associated with the likelihood of requiring rescue were assessed through multivariate logistic regression. Baseline patient characteristics ( Table 1 ) with a univariate association of p ≤0.25 were included in the full multivariate model (gender, weight, diabetes mellitus, previous myocardial infarction [MI], investigator-assessed MI location, investigator-assessed Q wave, and baseline creatine kinase-MB). Stepwise selection (p ≤0.25 as “in” criteria and p ≤0.05 as “out” criteria) was then used to identify the final factors; odds ratios and 95% confidence intervals (CIs) are reported for these factors.
| Variable | PI Strategy (n = 864) | Primary PCI (n = 927) | ||
|---|---|---|---|---|
| Rescue Angiography (n = 348) | Scheduled Angiography (n = 516) | p-Value | ||
| Age (years) | 60 (51–70) | 59 (50–67) | 0.163 | 59 (51–68) |
| ≥75 | 15.8% | 12.8% | 0.211 | 12.2% |
| Female sex | 16.4% | 20.9% | 0.095 | 21.9% |
| Weight (kg) | 80 (72–90) | 79 (70–88) | 0.012 | 80 (70–90) |
| Hypertension | 44.7% | 48.5% | 0.278 | 44.1% |
| Diabetes mellitus | 15.4% | 10.2% | 0.022 | 13.0% |
| Previous MI | 7.2% | 10.3% | 0.121 | 10.4% |
| Previous CHF | 0.0% | 0.6% | 0.278 | 1.6% |
| Previous PCI | 6.0% | 7.4% | 0.438 | 8.6% |
| Previous CABG | 0.0% | 0.4% | 0.518 | 0.3% |
| Heart rate (beats/min) | 75 (61–88) | 74 (62–85) | 0.425 | 75 (63–85) |
| Systolic blood pressure (mm Hg) | 135 (120–150) | 135 (120–150) | 0.766 | 137 (120–152) |
| Infarct location (investigator-assessed) | 0.040 | |||
| Anterior wall | 53.7% | 45.0% | 45.3% | |
| Inferior wall | 44.0% | 52.5% | 53.2% | |
| Other | 2.3% | 2.5% | 1.5% | |
| Killip class >I | 7.9% | 5.1% | 0.112 | 5.4% |
| Symptom onset to first medical contact (min) | 60 (39–101) | 61 (40–100) | 0.961 | 60 (35–98) |
| Symptom onset to baseline ECG (min) | 70 (45–115) | 73 (48–110) | 0.719 | 70 (42–107) |
| Symptom onset to randomization (min) | 88 (66–135) | 91 (68–130) | 0.728 | 91 (64–132) |
| Symptom onset to start of reperfusion treatment (min) | 99 (75–145) | 100 (75–140) | 0.578 | 177 (135–228) |
| Randomization to arrival at catheterization lab (min) | 128 (105–159) | 1086 (758–1314) | <0.001 | 65 (45–97) |
| Randomization to sheath insertion (min) | 140 (114–172) | 1101 (768–1333) | <0.001 | 76 (56–111) |
The association between reperfusion strategy received and clinical outcomes (i.e., the composite of all-cause death, cardiogenic shock, CHF, and reinfarction; and the individual components) at 30 days was examined using the Poisson regression model with robust error variance. Relative risks with 2-side 95% CI were reported, and these associations were adjusted for the TIMI risk score for STEMI.
All statistical tests were 2-sided with p value <0.05 considered as statistically significant. Statistical analyses were performed using SAS (version 9.3; SAS Institute, Cary, North Carolina).
Results
Figure 1 illustrates the derivation of the 3 patient cohorts that were evaluated in the present report. These consisted of (1) pharmacoinvasive patients with failed fibrinolysis who required rescue angiography (rescue, n = 348), (2) pharmacoinvasive patients with successful fibrinolysis who underwent scheduled angiography (scheduled angiography, n = 516), and (3) patients who were intended to undergo primary PCI (primary PCI, n = 927). Table 1 lists selected baseline characteristic in these 3 cohorts of patients. Rescue patients were more likely to be men, diabetic, have anterior MI, and tended to be men with Killip class >I than patients with scheduled angiography. After considering all patient characteristics in Table 1 , only MI location and body weight were significantly associated with the likelihood of rescue angiography (MI location: p = 0.038; anterior vs inferior MI: odds ratio 1.44, 95% CI 1.09 to 1.90; body weight [5-kg increase]: p = 0.009, odds ratio 1.06, 95% CI 1.02 to 1.11).
Table 1 also provides the key timelines for patients from symptom onset, randomization, and for acquisition of sequential 12-lead electrocardiograms. Overall, rescue patients underwent angiography promptly, that is, the median time from randomization was 140 minutes. Time from randomization to angiography was similar in those with an investigator-determined ECG protocol indication for rescue (i.e., <50% ST resolution, n = 258, median time 137 minutes [115 to 164]) and those with clinical indications (i.e., 13 for hemodynamic instability, 4 for refractory ventricular arrhythmias, and 73 were unspecified; median time 144 minutes [109 to 219]). For those undergoing scheduled angiography, the time from randomization to angiography was 1,101 minutes (median), that is, approximately 18 hours after TNK.
After TNK, 32 patients with successful reperfusion who were intended for scheduled angiography subsequently required urgent angiography within 6 hours of randomization because of new ECG, hemodynamic, or electrical evidence of ischemia as assessed by the local investigator ( Table A1 ). The median time from symptom onset to randomization was 95 minutes, whereas the interval between randomization and angiography was 217 minutes ( Table A2 ). Baseline ECG metrics were worst lead ST elevation of 3.0 mm, sum ST-segment deviation of 12.5 mm, and Q wave in infarcted territory of 31.3% ( Table A2 ).
Table 2 lists the angiographic findings and revascularization procedures for both patients receiving pharmacoinvasive therapy and those undergoing primary PCI. Although not statistically significant, patients who underwent rescue tended to more frequently have left anterior descending coronary artery as their culprit vessel than those who underwent scheduled angiography; there was no difference in the frequency of concomitant multivessel disease. At the time of the baseline angiogram, TIMI 2 or 3 epicardial coronary flow was present in 58.0% of patients undergoing rescue angiography, 88.0% in those undergoing scheduled angiography, and 29.6% in those undergoing primary PCI. Among patients who did undergo PCI, the subsequent TIMI 3 flow was lower in patients undergoing rescue and primary PCIs than in those undergoing scheduled angiography. Interestingly, after protocol-defined initial reperfusion therapy, medical management without the need for revascularization (i.e., neither PCI nor coronary artery bypass grafting [CABG]) was less frequent in patients undergoing rescue (11.5%) and primary PCI (7.3%) compared with those undergoing scheduled angiography (18.6%).
| Variable | PI Strategy (n = 864) | Primary PCI (n = 927) | ||
|---|---|---|---|---|
| Rescue Angiography (n = 348) | Scheduled Angiography (n = 516) | p-Value | ||
| Infarct-related coronary artery | ||||
| Left anterior descending | 49.9% | 43.4% | 0.065 | 45.4% |
| Left circumflex | 11.1% | 14.2% | 0.188 | 10.2% |
| Right | 37.9% | 41.6% | 0.282 | 44.4% |
| Left main | 1.2% | 1.4% | 0.766 | 0.4% |
| Normal coronaries | 0.6% | 1.8% | 0.127 | 0.7% |
| Multi-vessel disease | 45.7% | 49.3% | 0.305 | 45.4% |
| Baseline angiogram TIMI grade | <0.001 | |||
| 0/1 | 42.0% | 12.0% | 70.4% | |
| 2 | 18.1% | 13.9% | 10.1% | |
| 3 | 39.9% | 74.1% | 19.5% | |
| Post-angiogram/PCI TIMI grade | <0.001 | |||
| 0/1 | 5.8% | 2.2% | 4.1% | |
| 2 | 8.9% | 2.7% | 3.8% | |
| 3 | 85.3% | 95.1% | 92.2% | |
| % of patients receiving PCI | 86.5% | 76.2% | <0.001 | 91.5% |
| No. of patients receiving PCI (n) | 301 | 393 | 848 | |
| Stent placement during PCI | 93.0% | 98.0% | 0.001 | 95.3% |
| Only post-PCI TIMI grade, % | <0.001 | |||
| 0/1 | 5.4% | 1.1% | 2.8% | |
| 2 | 7.8% | 1.6% | 3.6% | |
| 3 | 86.8% | 97.3% | 93.6% | |
| Coronary artery bypass grafting | 4.3% | 5.4% | 0.459 | 2.2% |
Revascularization with CABG was more frequent in those receiving a pharmacoinvasive strategy than primary PCI, especially in those undergoing scheduled angiography ( Table 2 ). Among patients who underwent CABG during the index hospitalization, 53.3% (8 of 15) had previously undergone PCI in the rescue group, 45.0% (9 of 20) in the primary PCI group, and 3.6% (1 of 28) in the scheduled angiography group.
Table 3 lists the core laboratory ECG metrics. On the baseline electrocardiogram, rescue patients tended to have more Q waves and had significantly greater ST elevation in their worst lead and a greater sum of ST-segment deviation compared with patients undergoing scheduled angiography. After angiography and PCI, there was improvement in the ST-segment ECG metrics in the rescue cohort. Patients undergoing scheduled angiography had a more favorable ECG profile after treatment in comparison with the patients undergoing rescue PCI, and this was further enhanced after angiography and PCI.
| Variable | PI Strategy (n = 864) | Primary PCI (n = 927) | ||
|---|---|---|---|---|
| Rescue Angiography (n = 348) | Scheduled Angiography (n = 516) | p-Value | ||
| Baseline ECG | ||||
| Worst lead ST-E (mm) | 3.0 (2.0–5.0) | 2.5 (2.0–4.0) | <0.001 | 3.0 (2.0–4.0) |
| Sum ST-deviation (mm) | 15.0 (10.0–21.0) | 13.5 (9.0–19.5) | 0.004 | 13.5 (9.0–19.0) |
| Q-wave | 35.8% | 30.2% | 0.088 | 31.9% |
| Post-treatment ECG | 90-min Post tenecteplase | 30-min Post angiogram/PCI | ||
| Symptom onset to post-treatment ECG (min) | 181 (155–226) | 191 (164–235) | 0.022 | 260 (208–321) |
| Worst lead ST-E (mm) | 2.0 (1.0–3.5) | 0.5 (0.0–1.0) | <0.001 | 1.0 (0.0–1.5) |
| Sum ST-deviation (mm) | 11.0 (6.5–18.0) | 3.5 (1.5–5.8) | <0.001 | 4.0 (2.0–7.5) |
| Worst lead residual ST-E (mm) | <0.001 | |||
| <1 | 8.8% | 46.3% | 37.5% | |
| 1 to <2 | 24.5% | 39.0% | 37.7% | |
| ≥2 | 66.7% | 14.7% | 24.8% | |
| Worst lead ST-E resolution ≥50% | 32.7% | 89.0% | <0.001 | 82.8% |
| Post-angiogram/PCI ECG for PI strategy | ||||
| Symptom onset to post-angiogram/PCI ECG (min) | 325 (270–390) | 1291 (960–1520) | <0.001 | |
| Worst lead ST-E (mm) | 1.0 (0.0–1.5) | 0.5 (0.0–1.0) | <0.001 | |
| Sum ST-deviation (mm) | 4.5 (2.5–8.0) | 2.0 (1.0–4.0) | <0.001 | |
| Worst lead residual ST-E (mm) | <0.001 | |||
| <1 | 31.5% | 59.7% | ||
| 1 to <2 | 40.6% | 32.4% | ||
| ≥2 | 27.9 | 7.9 | ||
| Worst lead ST-E resolution ≥50% | 78.0% | 94.2% | <0.001 | |
| Post-PCI | ||||
| No. of patients receiving PCI | 301 | 393 | 848 | |
| Worst lead ST-E (mm) | 1.0 (0.0–1.5) | 0.5 (0.0–1.0) | <0.001 | 1.0 (0.0–1.5) |
| Sum ST-deviation (mm) | 4.5 (2.5–8.0) | 2.0 (1.0–4.0) | <0.001 | 4.0 (2.0–7.5) |
| Worst lead residual ST-E (mm) | <0.001 | |||
| <1 | 32.9% | 60.7% | 36.8% | |
| 1 to <2 | 38.3% | 31.8% | 38.0% | |
| ≥2 | 28.8% | 7.5% | 25.2% | |
| Worst lead ST-E resolution ≥50% | 78.1% | 93.5% | <0.001 | 83.8% |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
