The STrategic Reperfusion Early After Myocardial Infarction (STREAM) study investigated whether thrombolysis at first medical contact, followed by timely angiography or rescue percutaneous coronary intervention (PCI) if indicated in patients with acute ST-segment elevation myocardial infarction (STEMI) presenting within 3 h not able to undergo primary PCI within 60 min, is an appropriate and effective alternative reperfusion strategy (pharmaco-invasive) to PPCI.
The objective was to evaluate the outcome of patients presenting with STEMI within 3 h of symptom onset treated by the early administration of tenecteplase followed by angiography within 6-24 hours unless rescue intervention is indicated. The comparator arm consisted of the treatment with routine primary PCI alone according to local standards.
STREAM was an open-label, prospective, parallel, randomised, multinational trial to evaluate the outcome of acute STEMI patients presenting within 3 h of symptom onset, with approximately 1,000 patients in each arm. Patients who cannot undergo PCI within an hour are randomised to thrombolysis with tenecteplase within a pharmaco-invasive strategy or PPCI.1 It is important to note that STREAM was not a trial of lytic-facilitated PCI trial, in which all patients undergo immediate PCI.1 In STREAM, only lytic-treated patients with ECG or clinical evidence of failed reperfusion undergo immediate rescue intervention; the others undergo cardiac catheterisation (and any required follow-up procedures) between 6 and 24 h.1
Figure 1. STREAM: Study design
The primary endpoint at 30 days was a composite of death from any cause, shock, congestive heart failure (CHF), or re-infarction.1
1,915 patients from 15 countries were enrolled between March 2008 and July 2012. Of these, 1892 patients were randomised and provided informed consent to receive pharmaco-invasive therapy (n=944) or primary PCI (n=948).1
The primary endpoint occurred in 116 of 939 patients (12.4%) in the thrombolysis group and in 135 of 943 patients (14.3%) in the PPCI group (relative risk in the thrombolysis group, 0.86; 95% confidence interval [CI], 0.68 to 1.09; p=0.21).1
Figure 2. STREAM: primary endpoint at 30 days
Figure 2 shows the STREAM primary endpoint at 30 days.2 This result was consistent across the pre-specified subgroups, namely age, gender, Killip class, time to randomisation, place of randomisation, infarct location, systolic blood pressure, weight, history of diabetes or hypertension, TIMI risk score, and randomisation before or after the protocol amendment.2
When looking at the individual components of the primary endpoint, cardiogenic shock and congestive heart failure tended to occur more frequently in the primary PCI group than in the thrombolysis group (Figure 3).2
Figure 3. STREAM: Individual components of the primary endpoint at 30 days
The incidence of intracranial bleeding was higher in the pharmaco-invasive group compared to the PPCI group (1.0% vs. 0.2%, p=0.04), but this was reduced when the protocol was altered so that patients older than 75 years of age received half-dose tenecteplase instead of full-dose (intracranial bleeding: 0.5% for pharmaco-invasive group vs. 0.3% for PPCI, p=0.45).2
- A strategy of thrombolysis with bolus tenecteplase and contemporary antithrombotic therapy given before transfer to a PCI-capable hospital coupled with timely coronary angiography:1,2
- Circumvents the need for an urgent procedure in about two thirds of thrombolytic-treated STEMI patients.
- Is associated with a slightly increased risk of intracranial bleeding.
- Is as effective as PPCI in STEMI patients presenting within 3 hours of symptom onset who cannot undergo PPCI within one hour of first medical contact.
- STREAM has demonstrated how established fibrinolytic therapy has evolved to a new fibrinolytic strategy and how it can be safely and efficiently delivered. For patients with STEMI around the world who do not have timely access to PPCI, pharmaco-invasive treatment strategy should be implemented unless contraindicated, with fibrinolytic therapy followed by timely coronary angiography, if logistically and economically possible.3
This pre-specified analysis addresses the issues of increasing time delay of therapy on patient outcomes.4
Composite of death/congestive heart failure/cardiogenic shock/myocardial infarction in the pharmaco-invasive (PI) strategy and PPCI arms occurred in 10.6% vs 10.3% (≤55 min, p=0.910); 13.9% vs 17.9% (>55–97 min, p=0.148) and 13.5% vs 16.2% (>97 min, p=0.470), respectively.4
Outcomes of patients in the PI arms were consistent across different time groups, whereas outcomes worsened for patients in the PPCI arm as PCI-related delay increased (p(trend)=0.038). A PCI-related delay of 55 minutes or less was associated with fewer rates of adverse events in the PPCI arm. However, the PI strategy showed better outcomes when the time delay exceeded 55 minutes, suggesting beneficial effects of the PI strategy as PCI-related delay increases (p(interaction)=0.094).4
Figure 4. Relative association of continuous PCI-related delay (min) and pharmaco-invasive treatment with 30-day composite outcome (CHF, shock, MI)
[Gerschlick et al., 2015].4
The analysis demonstrated an increasing trend towards improved outcomes for PI patients with every 10-minute increment in delay (p(interaction)=0.073).4
The pharmaco-invasive strategy demonstrated a trend towards favourable outcomes compared to PPCI as PCI-related delay increased beyond guideline-mandated timelines. A pharmaco-invasive approach may therefore provide an alternative treatment option in cases of prolonged delays.4
Evaluated the pre-specified STREAM endpoint, aborted MI (AbMI), to compare a pharmaco-invasive strategy with PPCI.5
Rates of aborted MI (AbMI) were significantly higher in PI patients (11.1%) compared to in patients treated with PPCI (6.9%) (p<0.01). The likelihood of experiencing an AbMI was associated with fewer Q waves at baseline, less summed ST-deviation and shorter total ischaemic times.5
PI patients with AbMI had a significantly shorter time to fibrinolysis (90 vs. 100 min, p=0.015), while the total ischaemic time was 100 min longer in PPCI patients with AbMI. No ischaemic time difference was observed in PPCI patients with or without AbMI.5
No significant interaction was observed between rates of AbMI and treatment regarding the composite endpoint including death, shock, congestive heart failure and recurrent MI (p=0.292), but among PI patients, the incidence of this endpoint was significantly lower in patients experiencing an AbMI (5.1% vs. 12.0%; p=0.038). No such difference emerged in patients treated with PPCI.
A total of 45 patients (2.5%) balanced across treatment arms did not show ST-elevation but rather had infarct masquerade.
Results of the 30-day outcomes across the AbMI and non-AbMI groups irrespective of the treatment is presented in Figure 4.5
Table 1. Relative risk plot of the primary composite outcome and its components. AbMI vs non-AbMI
AbMI, aborted myocardial infarction; CHF, congestive heart failure; reMI, recurrent myocardial infarction
*Adjusted for Thrombolysis In Myocardial Infarction (TIMI) risk score
Continuous variables presented as median (25th-75th percentile)
*5.1 vs 12.0%, p=0.038
#The composite primary endpoint was death, cardiogenic shock, congestive heart failure, and recurrent myocardial infarction
Data from Maleki et al. 2014.5
A pharmaco-invasive strategy of early fibrinolysis coupled with anti-thrombotic and antiplatelet therapy more frequently aborts MI than PPCI. Such patients showed more favourable outcomes compared to non-AbMIs.
Comparison of ECG metrics and clinical outcomes in three groups; i.e. fibrinolysis requiring 37 rescue angiography (n=348), fibrinolysis with scheduled angiography (n=516), and 38 PPCI (n=927).6
Compared to those pharmaco-invasive patients undergoing scheduled angiography, rescue-fibrinolytic patients were more likely male, diabetic and have anterior MI. They also had more baseline sum ST-segment deviation and Q-waves.6
Thirty minutes post angiography, residual ST-segment elevation ≥2mm occurred in 27.9%, 7.9%, and 24.8% of rescue, scheduled and PPCI patients, respectively.6
Thirty-day composite event rates in these groups were: rescue 18.7%, scheduled 5.5%, and PPCI 13.9%; all-cause death: 6.1%, 2.1%, and 3.9%; shock: 7.5%, 2.0%, and 5.4%; heart failure: 11.8%, 2.3%, and 7.6% and recurrent MI: 1.5%, 1.4%, and 2.2%, respectively (Figure 5).6
Figure 5. Reperfusion strategy and 30-day composite endpoint and individual clinical events.
[Welsh et al., 2014].6
Compared to patients undergoing scheduled angiography post TNK, the adjusted relative risk (RR) of the primary outcome was 2.92 (95% confidence interval (CI) (1.92–4.45)) in patients undergoing rescue, and 2.32 (RR; 95% CI (1.57–3.44)) in patients treated with PPCI.6
Whereas early STEMI patients randomised to a pharmaco-invasive strategy had overall similar outcomes to PPCI, fibrinolytic-treated patients not requiring rescue had ECG evidence of superior reperfusion and lower clinical event rates than PPCI. By contrast those patients requiring rescue angiography had higher risk with more baseline co-morbidities and worse 30-day outcomes.6
Baseline characteristics, clinical outcomes, and the relationship of the TNK dose reduction to the efficacy, safety, and electrocardiographic indicators of reperfusion efficacy were evaluated in STEMI patients ≥75 years.7
Before TNK dose reduction, 3/42 of the elderly (≥75 years) patients (7.1%) suffered an intracranial haemorrhage (ICH), of which two cases were fatal.
No ICH occurred after dose amendment (half dose for patients ≥75 years) in 93 elderly patients receiving half-dose TNK.7
The numbers of elderly patients showing either a median extent of ST-segment resolution (≥50%) or those with ≥2 mm in the electrocardiogram lead with greatest ST-segment elevation were similar before and after TNK dose amendment (63.2% vs. 56% and 43.6% vs. 40.0%, respectively).7
Likewise, the proportion of patients requiring a rescue coronary intervention after fibrinolysis were also comparable pre- and post-amendment (42.9% vs 44.1%). Rates of the primary composite endpoint (30-day all-cause death, cardiogenic shock, congestive heart failure, and reinfarction) were 31.0% before amendment and 24.7% after amendment (Figure 6).7
Figure 6. Relative association of amendment and study treatment with 30-day primary composite end point in elderly patients.
Adjusted for TIMI risk score for STEMI. [adapted from Armstrong et al., 2015].7
In elderly STEMI patients, a half-dose of TNK reduces the likelihood of ICH without compromising efficacy of reperfusion. These observations are hypothesis generating and warrant further confirmation in randomised clinical trials in larger populations.7
The STREAM-2 study will compare outcomes in elderly patients (≥70 years) with acute STEMI randomised within 3 hours of onset of symptoms to a pharmaco-invasive strategy using half-dose tenecteplase followed by catheterisation within 6-24 h or rescue coronary intervention as necessary, or standard primary PCI.
For more details, click here
To determine the effect of the STREAM treatment strategies on 1-year mortality.8
Information on vital status at one year was available for 99.2% (936/944) of PI treated patients and 99.3% (941/948) of patients treated with PPCI.
Rates of all-cause mortality at one year were comparable for patients in both treatment categories (PI: 6.7% vs PPCI: 5.9%; p=0.49; risk ratio, 1.13; 95% confidence interval, 0.79-1.62) (Figure 7).8
Likewise, 1-year cardiac mortality was also similar in both treatment groups (PI: 4.0% vs PPCI: 4.1%; p=0.93; risk ratio, 1.13; 95% confidence interval, 0.62-1.54) (Figure 7).8
Figure 7. All-cause mortality / cardiac mortality after 1-year
<img data-cke-saved-src="/sites/default/files/07-All-cause-mortality.png" src="/sites/default/files/07-All-cause-mortality.png" width="600" "="" alt="All cause mortality Bar graph showing all cause mortality or cardiac mortality for STREAM study after 1 year" title="All-cause mortality / cardiac mortality after 1-year">
Between day 30 and 1 year, a total of 34 patients died (20 PI patients, 14 PPCI patients), 20 of 34 deaths were non-cardiac related (13 PI patients, 7 PPCI patients).
Overall, there was no significant difference in 1-year all-cause mortality between the 2 treatment groups among the pre-specified subgroups.
At 1 year, mortality rates in the PI strategy and PPCI arms were similar in STEMI patients presenting within 3 hours after symptom onset and unable to undergo PPCI within 1 hour.8
A pharmaco-invasive strategy for early presenting ST-segment elevation myocardial infarction nominally reduced 30-day cardiogenic shock and congestive heart failure compared with PPCI. This study evaluated whether infarct size (IS) was related to this finding. IS was divided into 3 groups: small (≤2 times the upper limit normal [ULN]), medium (>2 to ≤5 times the ULN) and large (>5 times the ULN).9
Patients of both treatment groups were stratified by IS. Overall, a higher proportion of patients treated with PPCI vs. PI showed large IS, despite similar ischaemic times between groups for each treatment strategy (PI vs. PPCI: small IS - 49.8% vs 50.2%; medium IS - 56.9% vs 43.1%; large IS - 48.4% vs 51.6%; p=0.035).9
An accumulation of shock and congestive heart failure was observed with increasing IS in both treatment arms with exception of the small IS group.
Within the small IS group, the difference in shock and congestive heart failure in favour of the PI strategy (4.4% vs 11.6%, p=0.026) was most likely linked to higher rates of aborted myocardial infarctions (72.7% vs 54.3%, p=0.005).
A trend for a beneficial effect favouring the PI strategy remained after adjustment (relative risk 0.40, 95% CI 0.15 to 1.06, p=0.064). In the medium and large IS groups, no treatment-relates differences were observed (Figure 8).9
Figure 8. Relative risk plot of the 30-day composite endpoint of shock/CHF. Infarct size groups and 30-day shock/CHF by treatment strategy.
*Adjusted for thrombolysis in myocardial infarction risk score. CHF indicates congestive heart failure; PI, pharmaco-invasive; PPCI, primary percutaneous coronary intervention; ULN, upper limit normal. [adpated from Shavadia et al., 2015].9
A PI strategy may potentially alter the pattern of IS after STEMI, resulting in higher rates of medium IS and lower rates of large IS compared with PPCI. Despite similar numbers of small infarcts, PI patients in this small IS group had higher rates of aborted myocardial infarctions and less 30-day shock and congestive heart failure.9
- Armstrong PW, et al. The Strategic Reperfusion Early After Myocardial Infarction (STREAM) study. Am Heart J 2010;160:30-35(e31).
- Armstrong PW, et al. Fibrinolysis or primary PCI in ST-segment elevation myocardial infarction. N Engl J Med 2013;368:1379-1387.
- Bates ER. Evolution from fibrinolytic therapy to a fibrinolytic strategy for patients with ST-segment-elevation myocardial infarction. Circulation. 2014;130(14):1133-1135.
- Gershlick A, et al. Impact of a pharmacoinvasive strategy when delays to primary PCI are prolonged. Heart 2015;101:692-698.
- Maleki ND, et al. Aborted myocardial infarction in ST-elevation myocardial infarction: insights from the STrategic Reperfusion Early After Myocardial infarction trial. Heart. 2014;100(19):1543-1549.
- Welsh RC, et al. Outcomes of a pharmacoinvasive strategy for successful versus failed fibrinolysis and primary percutaneous intervention in acute myocardial infarction (from the STrategic Reperfusion Early After Myocardial Infarction [STREAM] study). Am J Cardiol 2014;114(6):811-819.
- Armstrong PW, et al. Reduced dose tenecteplase and outcomes in elderly ST-segment elevation myocardial infarction patients: Insights from the Strategic Reperfusion Early After Myocardial infarction trial Am Heart J. 2015;169(6):890-898.
- Sinnaeve PR, et al. STEMI patients randomized to a pharmaco-invasive strategy or primary PCI: The STREAM 1-Year mortality follow-up. Circulation 2014;130(14):1139-1145.
- Shavadia J, et al. Infarct size, shock, and heart failure: does reperfusion strategy matter in early presenting patients with ST-segment elevation myocardial infarction? J Am Heart Assoc. 2015;4(8):e002049.