Prehospital fibrinolysis versus primary percutaneous coronary intervention in ST-elevation myocardial infarction: a systematic review and meta-analysis of randomized controlled trials

Primary percutaneous coronary intervention (PPCI) is considered the preferred reperfusion strategy in patients with ST-elevation myocardial infarction (STEMI), provided it can be performed expeditiously by an experienced team [1], based on studies comparing PPCI with in-hospital fibrinolysis (FL) [2]. International guidelines also underscore the objective of a total ischemic time <2 h in STEMI, which may not always be achievable if a PPCI strategy is chosen. Prehospital FL and direct transfer to a percutaneous coronary intervention (PCI)-capable center, which should be considered only when the estimated total ischemic time is >2 h after first medical contact in STEMI patients, is recommended over transfer for in-hospital FL when possible [3, 4].

Importantly, the relative benefit of PPCI over fibrinolytic therapy is time-dependent [5]. The benefit of PPCI over prehospital FL is not clear among patients managed early in the prehospital setting. The Comparison of primary Angioplasty and Pre-hospital fibrinolysis In acute Myocardial infarction (CAPTIM) trial [6] was the first large-scale trial comparing the two strategies. It showed that PPCI was not associated with lower mortality rates than prehospital FL. That study was terminated before reaching its target sample size, however. Researchers in the more recent Strategic Reperfusion Early after Myocardial Infarction (STREAM) trial [7] also reported similar rates of mortality between prehospital FL or PPCI. Both studies were undersized to assess a difference in mortality. In a real-life French nationwide registry of STEMI, prehospital FL was associated with reduced mortality in comparison with PPCI [8]. Hence, the benefit of allocating resources to developing proximity centers with 24/7 PCI facilities over prehospital FL and transfer to hub PCI centers may be questionable. The principal objective of the present systematic review and meta-analysis of randomized controlled trials was to compare prehospital FL and PPCI in terms of mortality.

Ultimately, three trials [6, 7, 9], comprising two long-term follow-up studies of the latter [10, 11] and a post hoc analysis of one trial [12], were selected for the meta-analysis. The review process is depicted in Fig. 1. The endpoints were collected at 30 days in the STREAM and CAPTIM studies and at 90 days in the Assessment of the Safety and Efficacy of a New Treatment Strategy with Percutaneous Coronary Intervention (ASSENT-4 PCI) study. The major characteristics of the patients of each study and the rates of the clinical endpoints are detailed in Table 1.

The meta-analysis showed consistently that the rates of short-term (30/90 days) death (RR 0.94, 95 % CI 0.67–1.31) (Fig. 2a) and cardiovascular death (RR 0.95, 95 % CI 0.64–1.4) (Fig. 2b) were similar between prehospital FL and PPCI. Similar results were also found for 1-year death (RR 1.01, 95 % CI 0.75–1.34) (Fig. 2c) and MI (RR 1.37, 95 % CI 0.84–2.21) (Fig. 2d) between the two strategies, but with nonsignificant but notable heterogeneity (I ² index 30.5 % and 34.6 %).

Consistently, compared with PPCI, the risk of cardiogenic shock was lower (RR 0.67, 95 % CI 0.48–0.95) (Fig. 2e), but the risks of stroke (RR 3.57, 95 % CI 1.39–9.17) (Fig. 2f) and hemorrhagic stroke (RR 4.37, 95 % CI 1.25–15.26) (Fig. 2 g) were higher, in association with prehospital FL, whereas the risk of ischemic stroke (RR 2.44, 95 % CI 0.71–8.47) (Fig. 2 h) did not significantly differ between the two strategies. Finally, the risk of major bleeding (RR 1.31, 95 % CI 0.96–1.78) (Fig. 2i) was similar between the two strategies, but with a high level of heterogeneity between the studies (I ² = 59.8 %).

Although the meta-analysis of the composite primary endpoints showed similar results between FL and PPCI (RR 0.99, 95 % CI 0.82–1.2), it should be considered only as exploratory because different endpoints composed the primary outcome of different studies. The sensitivity analyses by random-effects analysis (Fig. 2) and by excluding study by study (Additional file 1: Table S1) did not show any major change in the results of the principal analysis.

Our meta-analysis suggests that prehospital FL is associated with similar early and late mortality rates compared with PPCI in patients with STEMI managed early after symptom onset in the prehospital setting. Prehospital FL is also associated with a decreased risk of cardiogenic shock. However, it was consistently associated with an increased risk of stroke and hemorrhagic stroke.

In STEMI patients presenting in the first hours after symptom onset, PPCI is the recommended default reperfusion strategy if performed according to the recommended time limits [1]. Because most of these patients present to non-PPCI-capable hospitals, this reperfusion strategy may be a major challenge in many regions of the world [13]. Hence, recent trials have shown that prehospital FL appears to be a valuable alternative [6, 7, 9]. Our meta-analysis confirms that mortality rates and, less consistently, ischemic composite outcomes are similar between the two reperfusion strategies. This is concordant with real-life registries reporting comparable in-hospital [14] and even higher 1- and 5-year survival rates [8, 15] associated with prehospital FL compared with PPCI in similar populations. A major advantage of prehospital FL is the time gained from symptom onset to reperfusion (i.e., from symptom onset to start of reperfusion treatment with tenecteplase or alteplase in the FL group and first balloon inflation in the PPCI group), ranging from 60 to 78 minutes in the studies included in our analysis [6, 7, 9]. Such times are reported in countries where the density of centers with PPCI facilities is relatively high. Hence, it may be speculated that in many other regions, such times are even longer. Indeed, the time from symptom onset to reperfusion is of critical importance for myocardial salvage [16]. The use of a single-bolus fibrinolytic therapy as described in the STREAM and ASSENT-4 PCI trials [7, 9], but not in the CAPTIM trial [6], may be preferable because of its quick and easy administration. The use of continuous intravenous infusion of alteplase may also explain the relatively shorter difference of time from symptom onset to reperfusion therapy between the PPCI and FL groups in the CAPTIM trial [6] than in other trials.

Prehospital FL is recommended [1, 3] in association with early routine angiography, and PCI if needed. In the studies included in our analysis, patients were systematically transferred to PCI-capable centers and had high rates of PCI (70.4 %, 80.4 %, and 86.7 %, respectively, in the FL groups of the CAPTIM, STREAM, and ASSENT-4 PCI trials). The Which Early ST-elevation myocardial infarction Therapy (WEST) study [17] researchers previously showed that a strategy of early FL (only 18 % prehospital) coupled with routine early invasive management results in similar rates of death and recurrent MI compared with PPCI. Unlike these favorable results, the large Swedish registry [18] reported lower mortality rates in association with PPCI than both in-hospital and prehospital FL. Such discordance is probably explained by the low rates of PCI in the prehospital FL group (47.3 % of PCI or coronary artery bypass graft [CABG] within 14 days) in this registry. The optimal delay between successful FL and PCI is also a critical issue. Compared with PPCI, prehospital tenecteplase-facilitated PCI was associated with increased rates of reinfarction and trends toward higher rates of mortality [9] and increased infarct size [19]. Such results are explained by the paradoxical FL-induced platelet hyperreactivity and thrombin-induced platelet activation within the early hours after FL [20]. Hence, the time window of 3–24 h after successful FL is recommended [1] and explains the excellent results of the pharmacoinvasive strategy, as demonstrated in our meta-analysis.

Trends toward lower rates of cardiogenic shock occurrence with prehospital FL were noted in the CAPTIM [6] and STREAM [7] trials as well as in the WEST study [17]. Our meta-analysis shows a consistent and significant reduction of rates of cardiogenic shock complicating STEMI in association with prehospital FL compared with PPCI. This is also consistent with the fact that the most frequent cause of death was cardiogenic shock in the trials [6, 9]. In the CAPTIM trial [6], all shocks that occurred during the transport to the hospital were observed in the PPCI group, suggesting that the 1-h additional delay in the PPCI compared with prehospital FL may be responsible for the higher cardiogenic shock rates.

However, our meta-analysis highlights the consistent relative increase of the risks of stroke and hemorrhagic stroke in association with prehospital FL compared with PPCI. This finding may be tempered by the low absolute rates of stroke in each study, however, potentially explained by the limited proportion of elderly patients included. The increased risk of intracranial bleeding in patients ≥75 years old in the STREAM study [7] led to a dose reduction of tenecteplase in such patients. After the subsequent protocol amendment, no cases of intracranial hemorrhage were reported, and rates of mortality significantly dropped in the FL arm [11]. Moreover, rates of major bleeding did not significantly differ between the two strategies in our analysis.

The time from symptom onset to first medical contact as well as patient characteristics should be taken into account before selecting the reperfusion strategy [3]. In the studies selected in our analysis, patients were included if they presented within 3–6 h of symptom onset. However, the benefit of prehospital FL may be even more important in the first 2 h after symptom onset, as it has previously been reported to be associated with an improved 1-year survival compared with PPCI (2.8 % versus 6.9 %, respectively; p = 0.021) and low rates of intracranial hemorrhage (0.3 %) [21]. The patient’s risk profile, stratified by age and infarct location (anterior versus other), is also of importance in selecting the reperfusion strategy. In a large North American registry including 192,509 patients [5], the mean PCI-related time delay compared with FL, where mortality rates of the two reperfusion strategies were comparable, was calculated as 114 minutes in general. This delay was reduced to only 40 minutes in patients <65 years old presenting with an anterior MI within 2 h of symptom onset. In such patients with an extensive area of myocardium at risk, low risk of intracranial hemorrhage, and high chance of successful FL [22], a rapid and efficient restoration of flow with prehospital FL may improve myocardial salvage and outcome.

Our analysis suggests that prehospital FL is associated with similar early and late mortality rates compared with PPCI in patients with STEMI managed early after symptom onset in the prehospital setting. Prehospital FL was associated with a decreased risk of cardiogenic shock but an increased risk of stroke. Prehospital FL appears to be a valuable alternative to PPCI. Pharmacoinvasive strategies including prehospital FL and transfer to hub PCI centers may allow reduction of the cost allocated to developing proximity centers with 24/7 PCI facilities while providing similarly efficient reperfusion therapy.