Comparative effectiveness of primary PCI versus fibrinolytic therapy for ST elevation myocardial infarction: a review of the literature

A review of published literature on PubMed, including randomized controlled trials (RCTs) and observational studies published from 1 January 1985 to 1 September 2015. The following search terms were used: STEMI, myocardial infarction, fibrinolysis, percutaneous coronary intervention, PCI and angioplasty. Data were included if mortality or re-infarction was compared between the two reperfusion strategies. We also reviewed the current American Heart Association, European Society of Cardiology and NICE guidelines.

We reviewed a total of 14 RCTs, one meta analysis of RCTs and four observational studies. Two were published in the 1980’s, two in the 1990's and the remainder after 2000. Sample sizes for the RCTs ranged from 212 to 1892 covering a total of 10,221 patients. The meta-analysis included information for 194,040 patients. All but one of the studies included mortality in the composite of the primary end point, but with follow-up of varying durations following STEMI, between 30 days and 5 years. The reporting of other outcomes varied from trial to trial, including re-infarction, cardiogenic shock, arrhythmias, stroke, recurrent ischemia and congestive cardiac failure. Observational studies included a range of countries and cultures, where geographical barriers precluded many patients with STEMI from receiving pPCI within 120 min of first medical contact.

All present-day guidelines are in agreement regarding reperfusion strategies for the acute management of STEMI. That is, pPCI is the preferred treatment option if the onset of the patient's symptoms is less than 12 h and they have the ability to reach a facility with the required expertise within 120 min of first medical contact [1,2,12]. The American Heart Association quotes reduced 6 week (p = 0.0002) and long-term (p ≥ 0.0001) mortality rates among those who received pPCI compared with FL [12] Following FL, all guidelines recommend that patients receive early coronary angiography and, if appropriate, PCI.

Table 1 summarizes the results of the RCTs which have compared reperfusion strategies for STEMI over the last 22 years. In the 1990’s, the PAMI [6] and GUSTO IIb [7] trials both concluded a statisically significant reduction in mortality, reinfarction, (and in GUSTO IIb, stroke), 5.1% vs 12% (p = 0.02 [6]) and 13.6% vs 9.6% (p = 0.033 [7]), respectively, in those treated with pPCI. This remained the case even if pPCI was performed in a hospital without on-site cardiac surgery (C-PORT; 12.4% vs 19.9%; p = 0.03 [13]) and accounted for delays in treatment owing to transfer if the patient had presented to a hospital without the facilities to perform pPCI (DANAMI 2; 8% vs 13.7%; p = 0.002 [8]). While the negative prognostic effects of delayed transfer for pPCI were deemed minimal [14], it was noted that in the DANAMI trial, although patients were being transferred for pPCI, 96% were transferred in less than 120 min and therefore in fact complying with current guidelines.

Other trials were conducted to assess whether, after FL, outcomes were improved by subsequent PCI, regardless of clinical benefit post FL. The ASSENT-4 trial concluded an increase in 30-day mortality in those patients managed with PCI within 3 h following FL. This suggested a prothrombotic effect of fibrinolysis in early PCI and as a result subsequent trials delayed PCI until at least 3 h post FL [24]. Keeley et al. also demonstrated an increase in mortality, reinfaction, revascularization, bleeding and strokes in patients undergoing PCI post FL in a meta-analysis of 17 trials involving 4504 patients and as a result deemed it not suitable for use outside of the context of a RCT [25]. In contrast, CARESS in AMI [17], TRANSFER AMI [19] and NORDISTEMI [21] were all in agreement, that outcomes were improved if patients underwent immediate PCI following FL. CARESS in AMI analyzed a composite of death, re-infarction or refractory ischemia at 30 days (4.4% vs 10.7%; p = 0.004 [17]), similarly TRANSFER AMI assessed the same composite, but with the addition of congestive heart failure or shock (11% vs 17.2%; p = 0.004 [19]), with NORDISTEMI having a longer follow-up period of a year, with a composite of death, re-infarction and stroke (p = 0.01 [21]).

Further to this, RCTs then assessed outcomes according to the timing of PCI following FL. In 2006, the WEST [15] trial failed to demonstrate a significant difference in a composite of 30-day death, re-infarction, refractory ischemia, congestive heart failure, cardiogenic shock and major ventricular arrhythmia between patients who received pPCI, FL and immediate PCI or FL with standard care. In addition, there was still no difference for death and recurrent myocardial infarction between those undergoing pPCI and those managed with FL followed by immediate PCI. These results were echoed in the GRACIA-2 [16] trial in 2007, the 5-year follow-up to the CAPTIM [17] trial in 2009 and the STREAM [22] trial in 2013. In the STREAM trial, patients were randomized if unable to undergo pPCI within the next hour, therefore in keeping with the current guidelines. The recent FAST-MI [23] trial published in 2014 also concluded that there was no significant difference in all-cause mortality depending on whether pPCI or FL was the initial therapy. While a Cox multivariate analysis did show significant survival advantage in the FL group compared with those who underwent delayed pPCI (over 90 min following the first call), this was not confirmed in a propensity score matched analysis.

Given a number of trials with small sample sizes, Westerhout et al. combined the CAPTIM and WEST trial data, and with a larger sample size were able to conclude a survival benefit from FL if delivered within the first 2 h. Yet, if FL was administered after the first 2 h no significant difference was seen between the different modalities of reperfusion [26]. These studies were in contradiction to the results of the PAMI [6] and GUSTO IIb [7] trials, which were performed 20 years earlier favoring pPCI as the preferred modality of reperfusion in STEMI.

As many of these trials included a small sample size, the meta-analysis by Huynh et al. [20] in 2009 included all RCTs and observational trials up to 2008. Mortality outcomes in patients managed with pPCI and FL for STEMI were tabulated. The results were in agreement with the original trials, concluding a reduction in mortality both in the short term (34%) and long term (24%) in the pPCI group, therefore, continuing to recommend pPCI as the preferred management option.

Observational studies which attempted to compare outcomes from pPCI with those from FL differed. The WIRE registry [27] in Poland, 2008 found that only those STEMI who received early PCI had long term reductions in mortality. A conclusion echoed in the Middle East by Al-Zakwani et al. [28] who demonstrated that those treated with pPCI were less likely to have recurrent ischemic events, (odds ratio: 0.18; 95% CI: 0.06–0.56; p = 0.003). The STEPP-AMI study was a prospective, observational pilot study conducted in India that failed to show any difference in outcomes among those treated with pPCI or FL [29]. The primary end point was a composite of death, cardiogenic shock, re-infarction, repeat revascularization of a culprit artery and congestive heart failure at 30 days (11.1% vs 3.9%; p = 0.07). An Italian study by Manari et al. [30], however, concluded that despite longer door-to-balloon times for patients transferred from peripheral hospitals, delayed PCI was associated with favorable short and long-term outcomes.

Through the review of published trials, a meta-analysis and observational studies which included cohort from urban and a range of rural areas, we report the comparisons of the effectiveness of pPCI and FL for the acute management of STEMI. While pPCI remains the preferred management option for STEMI [1,2,12,18,20–22], more recent trials have confirmed that in the event of delayed PCI due to transfer times, that FL followed by PCI prior to discharge from hospital provided similar outcomes and is a viable option [16,17,19,23,27]. This finding appears more pertinent to STEMI arising from geographically remote areas, with the current guidelines reflecting this [1,2,12]. As the targeted reperfusion time is currently 120 min from first contact, the STREAM trial randomized patients if they were unable to receive PPCI within the next hour.

Although the more recent trials including STREAM, FAST-MI, GRACIA-2 and WEST [15,16,22,23] have all concluded mortality outcomes in favor of the FL group, it is important to note that none of the results were statistically significant and, therefore, while reperfusion with FL has been demonstrated to have similar outcomes as pPCI, it is not known to have superiority. Westerhout et al., however, did demonstrate reduced mortality in the FL group but only if delivered within 2 h (p = 0.021) [26].

The relevance of these newer trials to clinical practice mainly lies in the heightened recognition of the possibility of early use of FL as the reperfusion strategy for STEMI which, notably, many junior doctors are no longer comfortable in managing independently. This is likely as a result of the current default position of reperfusion for STEMI in many modern healthcare systems (such as UK) being emergency angiography and PCI [31]. Consequent to the introduction of pPCI centers throughout UK, this resulted in 98.5% of STEMI in England and Wales receiving pPCI in 2014, but with wide between-center variation in the adoption of the service [11,32]. Although such a high proportion of patients with STEMI receive pPCI, there are can be delays to reperfusion associated with inter-hospital transfer. The GRACIA-2 and WEST trials concluded that this may in fact negate the benefit of pPCI over immediate FL [15,16]. Therefore, improving confidence in administering FL when the door-to-balloon time is >120 min will subsequently reduce the time to reperfusion and improve outcomes. Delays in reperfusion are known to increase the rates of morbidity and mortality [18] as half of potentially salvageable myocardium is lost in the first hour, and two-thirds lost within 3 h [1]. Hence, the greatest benefit gained from reperfusion therapy occurs within the first 2–3 h of symptom onset. Owing to the delays, however, in time from symptom onset to first medical contact, the current European Society of Cardiology guidelines advise reperfusion with the first 2 h following the first medical contact if the symptom onset is under 12 h.

Other clinical impacts of this research on patient care include the importance of adverse events which may arise from both modes of reperfusion. Reassuringly, in the most recent data, the STREAM trial noted that if a reduced dose of Tenecteplase is used in those over 75 years that there was no statistically significant difference in intracranial hemorrhage between those who received FL and pPCI [22]. Although not included in our aims, it must be considered too, that if a patient undergoes FL prior to PCI, they become at risk of exposure to adverse events from both modes of reperfusion.

Both demographic and supply factors (such as medical staffing and hospital capacity) are known to be associated with regional variation in pPCI rates [9,10]. Owing to its small area and easily accessible pPCI centers, pPCI is likely to remain the preferred treatment option for those in UK with acute STEMI. For those living in more rural and remote areas, the newer trials have provided reassurance that management with FL, prior to transfer to a PCI-capable center will result in similar outcomes to those who live in cities where there is easy access to pPCI facilities. On the basis of these study results, India, which has the highest burden of acute coronary syndrome in the world [33], has developed a national STEMI program. This incorporates both pPCI and FL followed by coronary angiography (with PCI, if required) within 24 h of first medical contact, depending on location within the country [29]. This aims to improve overall outcomes in India as currently, among adults of working age, nearly 18 million productive years of life are expected to be lost from coronary artery disease by 2030. Remarkably, this is nine times higher than that expected in the USA within the same time-frame [34].

It is possible to debate how pPCI services in UK will evolve, by comparing our care to that delivered in other countries. Sweden for example, has more complete use of evidence-based practice [35], with new technologies introduced at an accelerated rate [36]. As a result, in Sweden, there is much lower hospital variation (interquartile range 16.7% vs 50.7%) [37] in guideline recommended treatment for acute myocardial infarction. In turn, this was associated with a higher 30-day mortality in the UK [37,38], and greater variation in that mortality [37]. A reduction in treatment variation has been associated with the number of physicians and nurses per 100,000 inhabitants, population density and the number of hospital beds per 100,000 inhabitants [9,10].

Thus, to develop a progressive pPCI service, more centers may be required in order to meet demand. However, this will increase staffing requirements, with a greater emphasis on experienced operators working out of regular hours. Alternatively, pPCI may become even more centralized, with larger centers responsible for intervention and perhaps patients undergoing FL at their nearest hospital prior to transfer at a later date. A study by West et al. [39] published in 2011 investigated outcomes based on the volume and the proportion of cases of STEMI who received pPCI. The authors demonstrated that although there was no significant difference in mortality depending on the volume of cases undertaken, that in centers where of all the STEMI patients, a smaller proportion (<25%) were treated with pPCI, there was an increase in mortality. In fact, those with a higher volume of cases had a shorter door-to-balloon times. This study therefore, promoted the use of fewer, larger centers.

Finally, any potential changes in the guidelines regarding reperfusion for STEMI will need to assess the financial implications, as increasing the number of patients receiving both FL and subsequent PCI will result in increased costs for the modern healthcare systems. In 2005, Selmer et al. [40] conducted a state-transition model that followed patients from the day of STEMI until death, with lifetime costs of €19,250 for pPCI, €24,000 for those undergoing pPCI that required transport from a peripheral hospital and €29,250 for FL. For those treated with FL but subsequent coronary angiography as now recommended in the guidelines, the NHS quote a cost of extra £1284 per patient, with a rise to £3837 for angioplasty (higher if more than one stent required) [41]. While we are unable to formulate comparisons between countries regarding cost of STEMI management as published data is from varying years and with different parameters used, the evidence does largely favor PPCI in Sweden (US$25,315 vs US$27,819) [42], USA (US$24,900 vs US$28,600) [43] and China (p < 0.05) [44]. The ZWOLLE trial, however, did conclude that FL while cheaper in the acute setting, by 1 year the difference has resolved (US$16,681 vs US$17,316) [45]. There are limited data from less developed countries regarding the cost effectiveness of the reperfusion options for STEMI management.

While pPCI remains the preferred management option for STEMI, more recent trials have confirmed that in the event of delayed PCI due to transfer times, that FL followed by PCI prior to discharge provided similar outcomes and is a viable option, particularly for those living in rural areas. The current guidelines reflect the available research, yet it is reassuring that in the event of a decline in pPCI services in the future that there remains a viable and evidence based alternative that will result in rates of mortality, re-infarction, recurrent ischemia, cardiogenic shock and heart failure similar to that of pPCI now, albeit at a higher cost.

In order to gain further evaluation of the two treatment options, a meta analysis including the newer trials would provide a more conclusive understanding of their clinical effectiveness.

Over the last ten years, the percentage of patients undergoing pPCI as the mode of reperfusion in STEMI management has increased and this is likely to continue. While pPCI remains the preferred treatment option, the increase in demand will need addressing by diffusion of services and increasing the number of centres providing a 24-h pPCI.

In order to improve patient outcomes further, we should aim to reduce the time between symptom onset and reperfusion by improving patient health behaviours and education. When in spite of this, geographical limitations prevent timely access to pPCI centres, there will always remain a role for FL in management of STEMI and the more recent studies have demonstrated that is a viable option for reperfusion with similar outcomes.

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.