• 
• 

Three Part Question

In [patients with acute ST elevation myocardial infarction] does [intravenous reteplase or intravenous streptokinase administration] lead to [lower mortality and superior probability of reperfusion]?

Clinical Scenario

A seventy-five year-old man with no significant previous medical history presents to the Emergency Department during the night with a two-hour history of typical cardiac chest pain. ECG demonstrates 2mm ST elevation in leads II, III and aVF with reciprocal ST depression in aVL. You diagnose acute inferior ST elevation myocardial infarction (STEMI).
Primary angioplasty is currently unavailable and you obtain verbal consent for intravenous thrombolysis. You wonder whether the more fibrin-specific bolus thrombolytic, reteplase, confers any advantage over streptokinase, in terms of mortality and probability of reperfusion.

Search Strategy

OVID Medline 1966 - 2006 August Week 3
OVID Embase 1980 - 2006 Week 33
OVID CINAHL 1982 to August Week 3 2006
The Cochrane Library 2006 Issue 3
Medline and Embase:
[exp Myocardial Infarction/ OR exp Coronary Thrombosis/ OR ((myocard$ adj infarct$) OR heart attack$).mp. AND [reteplase.mp. OR rapilysin.mp.] AND [exp Streptokinase/ OR streptokinase.mp. OR streptase.mp.] limit to humans and English language
CINAHL:
reteplase.mp. OR rapilysin.mp.
Cochrane:
[MeSH Streptokinase OR streptokinase (All fields)] AND [reteplase (All fields)]

Search Outcome

Medline: 54 papers, 2 relevant
Embase: 392 papers, 2 relevant
CINAHL: 33 papers, none relevant
Cochrane: 4 systematic reviews, 4 other reviews, 13 clinical trials and 2 economic evaluations. One paper was relevant to the three-part question (INJECT).

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
INJECT investigators 1995 Multinational	6010 patients from 208 centres in nine countries with symptoms and ECG criteria consistent with acute MI randomised to receive either streptokinase 1.5MU IV over 60 min or reteplase two boluses 10MU given 30min apart. All patients received placebo with double-dummy procedure; all patients received IV heparin	PRCT	35-day mortality	9.02% reteplase v. 9.53% streptokinase (90% CI -1.74%-0.73%). ARR 0.51% in favour of reteplase (NNT 196)	No explanation for selection of centres/countries involved. Only 90% confidence intervals stated for 35-day mortality (no explanation) Meta-analysis suggested that thrombolysis led to at least 2.1% mortality benefit (90% CI). The investigators therefore infer that if reteplase provided <=2.1% mortality benefit compared to streptokinase it was at least equivalent. This assumption is flawed. No robust definition for AMI – enzyme levels "defined locally" OR unequivocal new Q waves taken as evidence of AMI (doesn't fit the revised WHO criteria) "Enzyme levels" used as gold standard for AMI diagnosis – may not have been troponins Underpowered to detect small mortality benefits with reteplase (<1%)
			6-month mortality	11.02% reteplase v. 12.05% streptokinase. ARR 1.03% (95% CI -2.65% - 0.59%, p=0.217)
			Intracerebral events	37 (1.23%) v strep 30 (1.00%) ("non-significant")
			Death or continuing disability from in-hospital stroke at 35 days	9.19% reteplase v. 9.79% streptokinase (ARR -0.61%, 95% CI -2.09% - 0.88%)
			Recurrent MI or extension	No significant differences
			Subgroup analysis	Only significant differences were: Reteplase better if previous infarct; streptokinase better if systolic BP > 160mmHg (exact CI's and P values not given)
			Haemorrhagic strokes	Reteplase 23 (0.77%) v. streptokinase 11 (0.37%). Calculated odds ratio 2.09. No P value given.
Zeymer et al 1999 Germany	1154 patients who received thrombolytics, of whom 194 patients were treated with reteplase and 215 with streptokinase	Retrospective analysis of data from 6 angiographic trials comparing thrombolytic agents	90-minute patency of infarct related artery (assessed angiographically)	r-PA 76.3% TIMI 2 flow, 63.4% TIMI 3 flow; Streptokinase 59.3% TIMI 2 flow, 33.5% TIMI 3 flow. No direct statistical comparison	Retrospective analysis combining different trials – biased and unreliable comparison No baseline characteristics given Direct statistical comparisons not given In-hospital death is a convenient outcome but less robust than mortality at a specified time and subject to multiple biases Subgroup analysis only for mortality As this trial does not directly compare reteplase with streptokinase, it is impossible to draw meaningful conclusions regarding the relative efficacy of the agents. The data do suggest that there may be statistically significant improvements to thrombolytic efficacy with the fibrin-specific agents (reteplase and r-tPA) versus the non-fibrin specific agents (streptokinase, anysloloated plasminogen streptokinase activator complex and urokinase), especially >180 minutes after symptom onset.
			In-hospital death	Treated within 180 min of symptom onset: Reteplase 3% v. Streptokinase 4.5% (no direct statistical comparison); Treated >180min since symptom onset: Reteplase 1.6% v. streptokinase 14.5% (no direct statistical comparison)

Comment(s)

The stated aims of the INJECT trial were to prove that reteplase and streptokinase had equivalent thrombolytic efficacy in acute myocardial infarction, predominantly because of the large sample size necessary to demonstrate a small but statistically significant difference between the treatments. The investigators aimed to prove that reteplase yielded a 35-day mortality that is no more than 1% worse than streptokinase. They also state that meta-analysis suggests that the lower 95% confidence interval for the mortality benefit of thrombolysis over placebo is 2.1%. Given that a 1% difference between thrombolytic agents would potentially negate half of the proven benefit of thrombolysis over placebo, it is controversial to state that a 1% difference proves equivalence. Over 6000 patients were recruited to the INJECT trial but there were several significant weaknesses. Although the important outcomes of mortality at 35 days and 6 months were assessed, no attempt was made to assess the surrogate but potentially clinically important outcome of reperfusion (either angiographically or via ST segment resolution and clinical criteria). This may have been for logistic reasons, including the fact that the trial was multinational. Adequate collection of sufficient data may have been challenging. It is not immediately apparent why the authors stated only 90% confidence intervals for 35-day mortality. Assuming normality of the data, the upper limit of the 95% confidence interval for 35-day mortality would be 0.97%, as calculated by the author of this review. The second trial that was relevant to the three-part question involved a retrospective analysis of data from six angiographic trials, which randomised patients to various thrombolytic agents. By combining these data the research becomes subject to overwhelming potential bias. Although it appears that reteplase confers a mortality benefit in this analysis, the research is far from conclusive for the aforementioned reasons. The available evidence suggests that reteplase gives a 35-day mortality rate that is probably no more than 1% worse tahn streptokinase. On the balance of probability, they are at least equivalent, although the results of this trial are not conclusive. Contrary to traditional advice that plasminogen activators should only be considered in patients aged <75 years with anterior MI of less than 6 hours duration, basic science data suggests that reteplase may have superior thrombolytic efficacy when presentation is delayed (Zeymer et al, 1999). As the evidence is far from being unequivocally conclusive, it is important to take into account cost-effectiveness and a further short-cut review should be undertaken to assess this. Streptokinase costs £80-90 per patient, whereas reteplase costs £716 per patient (both excluding VAT). In practice within the United Kingdom, the convenience of bolus administration and the importance of fast thrombolysis and achievement of national targets may also be taken into account.

Clinical Bottom Line

The superiority and even equivalence of reteplase with streptokinase is not proven beyond reasonable doubt within the literature. The cost-effectiveness of reteplase over streptokinase should be assessed.

Level of Evidence

Level 2 - Studies considered were neither 1 or 3.

References

1. International Joint Efficacy Comparison of Thrombolytics Randomised, double-blind comparison of reteplase double-bolus administration with streptokinase in acute myocardial infarction (INJECT): trial to investigate equivalence Lancet 1995; 346: 329-336
2. Zeymer U; Tebbe U; von Essen R; Haarmann W; Neuhaus KL Influence of time to treatment on early infarct-related artery patency after different thrombolytic regimens American Heart Journal 1999; 137: 34-38