On August 17th 2013 The Lancet published what appears to be Eli Lilly’s face saving article on its derailed quest to replace its predecessor, clopidogrel (1).In its attempt to emulate Bristol Meyer with its triple C (CURE, CREDO, CLARITY) trials, pull the wool over the healthcare industry’s global eyes and become the second leading sold drug in the world, Eli Lilly hit a significant speed bump with the TRILOGY trial. Published in the NEJM in October of 2012, the trial found there was no difference in the composite endpoint of death, MI, or stroke in long-term use of prasugrel vs clopidogrel in patients with NSTEMI or unstable angina (2).Eli Lilly obviously hoped that this would be their CURE so to speak (3). Following the TRITON trial which showed questionable, albeit statistically significant benefit to the use of prasugrel for patients undergoing PCI (4), TRILOGY was meant to show this benefit in patients managed medically. To what I am sure was Eli LIlly’s utter dismay this was not the case.
But all was not lost. In an apparently pre-specified subgroup analysis that was not mentioned in their preliminary report (5) they found a significant improvement in their primary endpoint. Specifically in patients who underwent angiography to grade the extent of their plaque burden patients on prasugrel had statistically fewer heart attacks than those on clopidogrel. There was a 4% absolute difference in their primary composite endpoint of death, myocardial infarction and stroke that was almost completely due to a difference in the rate of myocardial infarction. However, as is becoming increasingly clear, catheterization for patients not having a STEMI leads to worse outcomes than medical management alone (6). So when NSTEMI and unstable angina patients are mismanaged initially, it appears that starting them on prasugrel results in fewer MI than clopidogrel. The unwritten conclusion of this trial is that since those patients who underwent angiography had lower rates of death and myocardial infarction, angiography to grade and correct any stentable lesions is the treatment of choice despite the multitude of randomized controlled trials showing the opposite (6). The authors state in their conclusion that when angiography is performed (unnecessarily) the patient will benefit from a more aggressive anti-platelet regime.
We could discuss the inappropriate nature of presenting findings in subgroup analysis as anything more than hypothesis generating propositions, especially when the overall findings of the trial are negative. I think the more important question is what is the clinical importance of these findings in the big picture? As far back as the CURE trial thienopyridines have been found to be minimally beneficial over aspirin alone. Consistently they have failed to find a mortality benefit (other than a 0.6% benefit in STEMI patients not treated with PCI in the COMMIT trial (7)). Thanks to the triple endpoint of death, MI, and stroke, clopidogrel and now prasugrel have been able to claim superiority to aspirin alone in ACS patients. What has been consistent throughout almost every thienopyridines trial is a 1-3% absolute reduction in MI. Since 2000 when the ESC/ACCF Guidelines allowed the diagnosis of myocardial infarction to be made with a rise and fall in troponin alone (8) and the increasing sensitivity/decreasing specificity of our troponin assays the question is “how do we evaluate the outcome measure myocardial infarction?” Is non-fatal myocardial infarction still a clinically relevant outcome or is it merely a disease-oriented outcome with questionable clinical value?
We have an inherently prognostic understanding of cardiac biomarkers. The reason we measure them in the emergency room is because of their ability to help us predict future morbidity and mortality in patients with suspected ACS. In September of 2000 a joint committee of the European Society of Cardiology and the American College of Cardiology changed the definition of MI to include those with a rise and fall in troponin alone, which was used by the authors of the CURE trial. Under this new definition, approximately 20% more patients would be diagnosed with a myocardial infarction as compared with the prior definition (10).Overall the patients diagnosed exclusively by the new definition had a more benign hospital course(mortality 7.6% vs. 3.9%, CHF 11.2 vs. 5.9%) but 6 month outcomes ended up similar to those diagnosed by the old definition (mortality of 13 vs. 19) (10). MI diagnosed by troponin alone allows us to identify a group of patients at risk of increased mortality that would otherwise be missed using the old definition and for this reason possesses some diagnostic utility.How though does it translate to an outcome measure in clinical trials where the prognostic events that troponins are meant to predict have already occurred?
The theory behind identification of non-fatal myocardial infarction as a surrogate endpoint for poor prognostic outcomes is based upon the belief that patients who have such an infarct are found to have increased morbidity and mortality in the future. In both the CURE and the TRILIOGY trials this did not hold true.In the CURE trial patients were followed for one year while the TRILOGY trial the median duration of follow up was 17 months. During this period there was no difference in mortality between the active and control groups. In the CURE trial the difference of heart failure on echo between the groups was 0.7%. If there is no difference in mortality and no clinical difference in heart failure, just how important clinically is this difference in myocardial infarctions?
To answer this question we must examine what types of infarctions are thienopyridines preventing. In the TRILOGY trial only patients who underwent angiography showed any benefit. In CURE if you exclude the patients who underwent PCI the difference in the rate of myocardial infarction between clopidogrel and placebo was 1%(13). In the COMMIT trial done on 45,000 patients, the majority of which had a STEMI without PCI, the mortality difference was 0.6%, which was statistically significant due to the large sample size, but has marginal clinical significance. The re-infarction rate was 2.4% in the placebo group vs. 2.1% in the clopidogrel group. Taken as a whole, these trials failed to find a benefit for patients who did not undergo PCI.
It seems reasonable to conclude that the PCI itself may be causing the myocardial damage that both clopidogrel and prasugrel alleviate. The TRITON trial can further elucidate this subject. The TRITON trial (4)compared aspirin and clopidogrel to aspirin and prasugrel in ASC patients scheduled for emergent/urgent PCI. It was a trial of over 12,000 patients, in which 26% were STEMIs and the remainder were NSTEMIs or unstable angina. As is typical with all the thienopyridine trials there was no difference in mortality between the two groups. The entire variation in the primary endpoint of death, MI and stroke was made up by a 2.5% difference in myocardial infarctions. The authors published a secondary analysis in which they examined the types of myocardial infarctions that occurred in their cohort (9). In this paper the authors use the ESC/ACCF guidelines and categorize the myocardial infarctions into 5 types. 1,2, and 3 are myocardial infarctions that are spontaneous in nature and 4 and 5 secondary to procedures (8). They also organized the infarctions by STEMI or NSTEMI. There were a total of 7.7% and 10.1% of infarcts in the prasugrel and clopidogrel groups respectively. This is more than what is accounted for in the original document but that seems to be because some patients had multiple MIs in different categories. In the spontaneous infarct category alone there is less than a one percent difference between the groups (2.8% vs. 3.7%).Whereas the majority of the difference is made up of procedural infarctions (4.9% VS 6.4%), accounting for 63.2% of the total infarctions in the entire cohort. Even the majority of STEMIs that occurred were secondary to PCI or CABG.
There have been numerous studies evaluating the clinical significance of cardiac biomarker elevation after PCI. Consistent elevations in CK-MB levels 5-8 times normal have been correlated with increased in-hospital stays and long term mortality (14). Unfortunately the same cannot be said for troponin. Studies examining the prognostic value of troponin in PCI-related myocardial damage have been frustratingly inconsistent. Some have found an increase in mortality with 3 times greater than normal level of troponin while others have found no association at all. A large meta analysis published in 2008 (11) found no consistent prognostic value of elevated troponins following PCI. An article published in Circulation in 2010 prospectively examining patients with isolated troponin elevations without concomitant elevations of CK-MB found no association between elevated troponins and mortality (16). In fact the clinical relevance of the majority of PCI related infarction is questionable at best and seem to be more an indicator of clot burden and procedural difficulties than long term morbidity and mortality (12). This is certainly the case in the TRITON trial since the small increase in PCI related myocardial infarction in the clopidogrel group did not translate into increased mortality at 15 months (9).
These small gains do not seem worth the excessive bleeding risk that is associated with thienopyridines. In the original CURE trial there was a 1% increase in overall bleeding, half of which was categorized as severe (3). In the TRITON trial there was a 1.2% absolute increase in bleeding with prasugrel when compared to clopidogrel, 80% of which required a blood transfusion (4). These bleeds were not severe enough to affect overall mortality, but the authors did not measure the long term morbidity that is associated with intracranial hemorrhage or intraocular hemorrhage.Both considered critical enough to be categorized as severe bleeding events in the CURE and TRITON trials(3,4).
PCI-TRILOGY and all the clopidogrel and subsequent prasugrel trials are a clear example of how composite endpoints taken at face value will demonstrate benefit where likely one does not exist. As we have become increasingly skeptical of the triple endpoint in cardiovascular trials so too should we be skeptical of its components. The mere fact that all infarctions are not created equal means we have to treat this single endpoint as a composite endpoint of its own. If a specific intervention reduces the rate of infarctions we must ask if these infarctions are worth the cost of the intervention both in medical resource utilization and adverse events secondary to that intervention.
- Wiviott, S et al. Prasugrel Verses Clopidogrel for Patients with Unstable Angina or Non-ST-Elevation Myocardial Infarction With or Without Angiography: A Secondary, Prespecified Analysis of the Trilogy ACS Trial. Lancet 2013; 382: 605-613
- Roe MT et al. Prasugrel Verses Clopidogrel with Acute Coronary Syndrome Without Revascularization. N. Eng J Med 2012; 367:1297-309
- Yusuf S et al. Effects of Clopidogrel in Addition to Aspirin in Patients with Acute Coronary Syndrome Without ST-segment Elevation. N Engl J Med 2001; 345:494-502
- Wiviott SD et al. Prasugrel Versus Clopidogrel in Patients with Acute Coronary Syndromes. N Engl J Med 2007; 357:2001-15
- Chin CT, Roe MT, Fox KA, et al. Study design and rationale of a comparison of prasugrel and clopidogrel in medically managed patients with unstable angina/ non-ST-segment elevation myocardial in-farction: the TaRgeted platelet Inhibition to cLarify the Optimal strategy to medi-cally manage Acute Coronary Syndromes (TRILOGY ACS) trial. Am Heart J 2010; 160(1):16.e1-22.e1.
- Stergiopoulus K et al. Initial Coronary Stent Implantation with Medical Therapy vs. Medical Therapy Alone for Stable Coronary Artery Disease. Arch Int Med, Vol. 172(NO.4) 2012
- COMMIT collaboration group. Addition of Clopidogrel to Aspirin in 45,852 Patients Acute Mycardial Infarction: Randomized Placebo-controlled Trial. Lancet 2005;366:1607-21
- Thygesen K, Alpert JS, Jaffe AS, Simoons ML, Chaitman BR, White HD; the Writing Group on behalf of the Joint ESC/ACCF/AHA/WHF Task Force for the Universal Definition of Myocardial Infarction. Third universal definition of myocardial infarction. Circulation 2012
- Morrow D et al. Effects of the Novel Thionopyridine Prasugrel Compared with Clopidogrel on Spontaneous and Procedural Mycardial Infarction in the Trial to Assess Improvement in Therapeuthic Outcomes by Optimizing Platelet Inhibition With Prasugrel- TIMI 38. Circulation 2009; 119:2758-2764
- Meier, M et al. The New Definition of Myocardial Infarction. Diagnostic and Prognostic Implications in Patients With Acute Coronary Syndromes.Arch Intern Med. 2002;162(14):1585-1589
- Nienhuis MB et al. Prognostic Value of Troponin After Elective Percutaneous Coronary Intervention: A Metanalysis. Catheter Cardiovasc Interv 2008;71:318-24
- Prasad A et al. Myocardial Infarction Due to Percutaneous Coronary Intervention. N Engl J Med 2011;364:453-64
- Mehta S et al. Effects of Pretreatment with Clopidogrel and Aspirin Followed by Long-Term Therapy in Patients Undergoing Percutaneous Coronary Intervention: The PCI-CURE Study. The Lancet, Vol 358 2001
- Stone GW et al. Differential Impact on survival of Electrocardiographic Q-wave Verses Enzymatic Myocardial Infarction After Percutaneous Intervention: A Device-Specific Analysis of 7147 Patients. Circulation 2001;104:642-7
- Christophe M et al. Patients with Acute Coronary Syndrome and Normal High-sensitivity Troponin. The American Journal of Medicine – December 2011 (Vol. 124, Issue 12, Pages 1151-1157.
- Cavallini, Claudio, et al. “Prognostic value of isolated troponin I elevation after percutaneous coronary intervention.” Circulation: Cardiovascular Interventions 3.5 (2010): 431-435.
Thanks for all your help!
- Peer reviewed by Brent Thoma
- Special thanks to Rebecca Talmud