- Email: [email protected]
Redesigning TRACER trial after TRITON
International Journal of Cardiology 197 (2015) 44–47
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
Editorial
Redesigning TRACER trial after TRITON Victor L. Serebruany ⁎ HeartDrug™ Research Laboratories, Johns Hopkins University, Osler Medical Building, 7600 Osler Drive, Suite 307, Towson, MD 21204, United States
a r t i c l e
i n f o
Article history: Received 16 May 2015 Received in revised form 13 June 2015 Accepted 16 June 2015 Available online 18 June 2015 Keywords: Clinical trials Trial design Vorapaxar Prasugrel Outcomes
a b s t r a c t Designing of smart clinical trials is critical for regulatory approval and future drug utilization. Importantly, trial design should be reconsidered if the interim analyses suggest unexpected harm, or conflicting results were yielded from the other trials within the same therapeutic area. With regard to antiplatelet agents, the perfect example is redesigning of the ongoing PRoFESS trial by eliminating aspirin from clopidogrel arm after the earlier MATCH trial results became available. The goal was to aseess the unchanged TRACER trial design in light of the evidence yielded from the earlier completed TRITON trial. TRACER was designed as a triple versus dual antiplatelet trial in NSTEMI patients with no previous long-term outcome data supporting such aggressive strategy. TRITON data represented dual versus dual antiplatelet therapy, and became available before TRACER enrollment starts revealing prasugrel front-loaded early vascular benefit predominantly in STEMI patients with the growing over time bleeding and cancer risks. Moreover, large prasugrel NSTEMI TRITON cohort exhibited trend towards excess mortality in experimental arm warning against aggressive TRACER design. The long-term TRITON results in general, and especially in the NSTEMI patients challenge unchanged TRACER trial design. Applying dual, rather than triple antiplatelet therapy protocol modification should be considered in TRACER to minimize bleeding, cancer, and non-cardiovascular death risks. © 2015 Published by Elsevier Ireland Ltd.
1. Introduction 1.1. Trial design Designing of clinical trials for regulatory drug assessment is a complex activity directly impacting chances for approval, future therapeutic options, and commercial success or failure. Importantly, trial designs are not written in stone, and are oftenly amended, adjusted, or even grossly changed. Usually, the main reasons leading to such dramatic trial redesign are unexpected findings discovered during interim analyses, or final results of other outcome-driven trials within the same therapeutic area. With regard to the antiplatelet agents, changes of then ongoing PRoFESS trial design are worth mentioning. Indeed, PRoFESS's initial objective sought to compare combination aspirin–dipyridamole to combination aspirin–clopidogrel [1]. However, after MATCH trial demonstrated that the addition of aspirin to clopidogrel had increased the risk of major bleeding complications without significantly reducing the risk of recurrent ischemic stroke [2], the PRoFESS protocol was subsequently modified. As a result, the 2027 patients initially randomized to receive aspirin–clopidogrel were taken off aspirin, and all the following patients randomized to that arm were administered clopidogrel monotherapy [3]. ⁎ Corresponding author at: 7600 Osler Drive, Suite 307 Yowson, MD 21204, USA. E-mail address: [email protected].
http://dx.doi.org/10.1016/j.ijcard.2015.06.020 0167-5273/© 2015 Published by Elsevier Ireland Ltd.
In short, trial design of ongoing active studies is a moving target, and it is the responsibility of trial leadership to watch closely for the constantly changing environment indirectly impacting expected clinical outcomes. Here, we will outline the comprehension of the Thrombin Receptor Antagonist for Clinical Event Reduction in Acute Coronary Syndrome (TRACER) trial [4] in light of the evidence yielded from the earlier TRial to assess Improvement in Therapeutic Outcomes by optimizing platelet InhibitioN with prasugrel Thrombolysis In Myocardial Infarction 38 (TRITON-TIMI-38) trial [5].
2. TRITON Landmark TRITON-TIMI 38 was a pivotal phase 3, randomized, double-blind, parallel-group, multinational, clinical study. The trial assessed head-to-head the efficacy and safety of the experimental antiplatelet agent prasugrel versus standard care with clopidogrel [5]. Patients (n = 13,608) with moderate to high-risk acute coronary syndromes (ACS) with the defined coronary anatomy undergoing percutaneous coronary intervention were randomized to prasugrel 60 mg loading dose followed by 10 mg daily or clopidogrel 300 mg loading dose followed by 75 mg daily for 6 to 15 months. The primary end point was the rate of cardiovascular (CV) death, myocardial infarction (MI), or stroke, and occurred in 12.1% of patients treated with clopidogrel, and 9.9% of patients randomized to prasugrel. This difference
Editorial
The trial randomized 12,944 patients with non-ST-elevation ACS in 37 countries throughout the world to receive vorapaxar (novel PAR-1 thrombin receptor antagonist) or placebo on top of standard care with clopidogrel and aspirin [4]. The symptom onset had to be within 24 h of presentation and eligible patients had to have either elevated troponins/creatinine kinase-myoglobin (CK-MB) or new ST-segment changes and at least one other high-risk marker such as prior MI or revascularization, peripheral vascular disease, or diabetes [4]. A composite of death from CV causes, MI, or stroke occurred in 822 patients in the vorapaxar group versus 910 in the placebo group (14.7% and 16.4%, respectively; hazard ratio, 0.89; 95% CI, 0.81 to 0.98; P = 0.02). Rates of moderate and severe bleeding were 7.2% in the vorapaxar group and 5.2% in the placebo group (hazard ratio, 1.35; 95% CI, 1.16 to 1.58;P b 0.001). Intracranial hemorrhage rates were 1.1% and 0.2%, respectively (hazard ratio, 3.39; 95% CI, 1.78 to 6.45; P b 0.001) in favor of placebo. Rates of nonhemorrhagic adverse events were similar in the two groups with the exception of excess of solid cancers after triple antiplatelet therapy [9]. Follow-up in the trial was terminated early after a safety review. 3.1. Chronology The time course of major activities for both trials is outlined in Table 1. The comparison of trial chronology suggests that the enrollment to TRACER starts after [4] the entire TRITON data became available [5]. Indeed, our comprehension of TRITON has been reevaluated after more balanced and detailed FDA reviews [6,7], than following primary overoptimistic publication [5]. However, all adverse but realistic TRITON findings were readily available before TRACER start, and could be carefully considered. 3.2. Target population Both trials share the dominant proportion of NSTEMI patients. In fact, over 94% in TRACER [4] and 74% in TRITON [5] experienced unstable angina, or non-STEMI ACS (NSTEMI) as qualifying event at enrollment. Since TRACER was designed specifically for NSTEMI patients, the outcome data, and especially TRITON mortality numbers in such target population are critically important, and should be considered before TRACER enrollment starts. Overall, in TRITON, while cardiovascular deaths slightly trended in favor of prasugrel (133 versus 150), however, excess in bleeding fatalities (21 versus 5), and 4 extra cancer-related deaths after prasugrel dilute the mortality benefit almost completely.
TRITON revealed few serious problems for prasugrel utilization as a chronic antiplatelet agent. In fact, there are 43 issues which should be considered for final TRACER design. First, despite the exclusion of heart surgery patients, those with high bleeding risks, and thrombocytopenia, prasugrel on top of aspirin cause highly significant increase of bleeding events than standard care with clopidogrel and aspirin combination [5–7]. In fact, the rates of key safety endpoint in TRITON, namely non-CABG major bleeding was 32% higher; spontaneous bleeds occur 52% more frequent; so as bleeding-related transfusions in the excess of 34% were observed in the prasugrel-treated patients. Moreover, fatal bleeding and major bleeding after CABG were over 4 times more often in the prasugrel arm. The second TRITON shortcoming was utilization of TIMI bleeding classification. This particular scale is the most conservative, focusing predominantly on major bleeding episodes, and excellently serving to assess safety of fibrinolytic agents. However, hemorrhagic strokes so prevalent with lytics are not common for modern antiplatelet strategies, when “insignificant” superficial bleeding episodes are most prevalent. By applying more liberal GUSTO bleeding classification, initial TRACER design will collect and report more bleeding events than filtered in TRITON. Since the TIMI scale-based existing TRITON bleeding rates after dual antiplatelet therapy were already known before TRACER start, applying triple antiplatelet therapy should be carefully comprehended for better and safer TRACER design. The third problem becomes obvious when considering the length of follow-up and bleeding risks in TRITON. Since the median follow-up in TRITON was 14.5 months, it was critical to realize when the most bleeding events occur. The FDA-generated Fig. 2 clearly answers this question with regard to the distribution of any bleeding events (A), and serious bleeding events (B). Indeed, there is an exponential growth of bleeding
Deaths in UA/NSTEMI Patients
0.00
Table 1 Chronology of TRITON and TRACER activities.
3.3. Bleeding
0.04
3. TRACER
Moreover, there were more deaths with unidentified causes after prasugrel (5 versus 2) in patients lost to follow-up without achieving primary endpoints, but affecting all-cause mortality [6,7]. In addition, the FDA provided a detailed distribution of deaths in TRITON, dependent on the type of qualifying initial occlusive event. The differences between mortality among patients with ST-elevated MI (STEMI), and UA/ NSTEMI in TRITON are presented in Fig. 1, and Table 1. There was a small but consistent trend (30 fatalities) towards more deaths in experimental TRITON arm among NSTEMI patients in contrast to STEMI cohort. This fact yielded from TRITON is a sufficient warning to be considered for TRACER design change.
fraction of patients 0.01 0.02 0.03
favoring experimental thienopyridine was predominantly attributed to the decrease of ischemic episodes during the first 3 days after intervention, and was especially evident in the STEMI cohort [6–8]. However, patients treated with prasugrel experienced higher bleeding rates, including significant increase of the key prespecified non-CABG-related TIMI major bleeding, and the reduced death from CV causes was canceled by the increased fatal bleeding on prasugrel [5–7], and surprising cancer signal after prasugrel [6,7].
45
Activity/trial
TRITON
TRACER
First patient in Last patient in Futility meeting MI definition changed Database locked Analyses completed Presentation/publication FDA reviews
November, 2004 January, 2007 July 2007 August 2007 September, 2007 October, 2007 November, 2007 February 2009
December, 2007 November, 2010 January, 2011 Not applicable September, 2011 October, 2011 November, 2011 April, 2014
0
2
4
6
Number at risk rx = Clopidogrel 5029 4867 4828 4735 rx = Prasugrel 5044 4896 4837 4742
8 month
10
12
14
4190 3562 3252 2767 4149 3542 3203 2695
rx = Clopidogrel
16 300 293
rx = Prasugrel
Fig. 1. Distribution of deaths over the follow-up in UA/NSTEMI patients in the TRITON trial. Exhibited from the FDA secondary review [6].
46
Editorial
Fig. 2. Distribution of any (A), and serious (B) bleeding events in TRITON. From the FDA reviews on prasugrel [7], and vorapaxar [9].
risks over time in the TRITON trial especially evident for serious bleedings. In fact, mean follow-up in TRACER was slightly longer (16 months) than in TRITON, increasing bleeding risks even further. Therefore, there is a compelling evidence that majority of bleeding events occur late, beyond the 30-days after initiation of therapy. This fact should have been kept in mind before finalizing TRACER protocol. Indeed, lack of direct comparisons, quoting bleeding incidence after prasugrel in TRITON does not negate the risk of bleeding with vorapaxar in TRACER. However, TRACER applied triple antiplatelet strategies, and if dual antiplatelet regiment in TRITON caused so much higher bleeding risks, then such obvious disadvantage should be kept in mind for potential modification of TRACER design after the TRITON data became available.
3.4. Cancer The excess of solid cancers after vorapaxar in TRACER (Hazard Ratio 1.4 (95% CI 1.1) top 1.8, p = 0.012) look very similar to those for prasugrel in TRITON [9] (see Fig. 3 for details). Importantly, the TRITON trial initiated the focus on the association of aggressive antiplatelet strategies, bleeding and cancer. In TRACER, as in TRITON, the recounted death numbers in all patients with solid cancer were high: 37 in vorapaxar patients vs. 24 in placebo [9]. As a rule of a thumb, more potent antiplatelet regimens cause more bleeding, and more
TRITON
solid cancers. Also, overall, the potency of antiplatelet therapy was high in TRACER, when triple antiplatelet therapy in the experimental arm (aspirin + clopidogrel + vorapaxar) exceeds 87% [4]. Finally, serious adverse events in TRACER were only to be reported until 60 days after the last dose of the study drug. While this limitation is less critical for adequate bleeding assessment, it is problematic for delayed event capturing, such as cancers, that take time to develop and be detected challenging triple antiplatelet strategies even further. Longer followup should be reinforced in future trials if we want to keep and assess delayed events such as cancers.
3.5. Impressions Indeed the trials have never been perfect, and every attempt should be made to identify the learning point from them in order to improve the next step. Obviously, there are limitations on extrapolation of the data from one trial to the other yet to start study. Indeed, any analogies may fail to work in practice and it is unclear that experience with a potent P2Y12 antagonist can be directly applied to a PAR-1 blocker. However, TRITON trial results in general, and especially in non-STEMI patients should warn against aggressive triple antiplatelet strategy later implemented in TRACER. The TRITON data were readily available before TRACER enrollment, but were not reconsidered resulting that the initial TRACER trial design has not been adjusted. The TRITON results
TRACER
Fig. 3. New solid cancers in TRITON and TRACER. From the FDA reviews on prasugrel [7], and vorapaxar [9].
Editorial
should impact TRACER protocol modification adding value to the increasing body of evidence suggesting that a more delicate strategy with modest platelet inhibition, especially in a chronic long-term setting, represents a substantial advantage over aggressive antiplatelet regimens. In this regard, applying triple antiplatelet strategies for the population with moderate ischemic burden (NSTEMI ACS), when excessive platelet inhibition is hardly justified was provocative, and contributes to the TRACER trial failure to gain NSTEMI indication. Ideally, TRACER should be designed with three arms: vorapaxar and aspirin; vorapaxar and clopidogrel; and versus standard care with clopidogrel and aspirin combination. If two arms will be the only practical option, then TRITON-type design directly comparing vorapaxar versus clopidogrel both on top of aspirin should apply. While distribution of thrombotic outcomes cannot be predicted, the bleeding and cancer signal will definitely look better for vorapaxar in TRACER. TRITON should not be overlooked for another important reason. Since clopidogrel substituted ticlopidine without any outcome-driven evidence, but rather on overcoming bone marrow toxicity safety concerns, TRITON was a first ever direct comparison of clinical outcomes between next generation thienopyridine and clopidogrel. If the TRACER leadership looks carefully at the TRITON results, they will realize that most if not all prasugrel benefit occur early, predominantly in STEMI cohort, with growing over time bleeding and cancer risks. The fact that all prasugrel patients at the end of TRITON were switched to clopidogrel reflects the stunning reality, and speaks for itself suggesting that TRITON leadership perfectly realize the escalating risks for keeping patients on prasugrel. Another issue is that when applying triple antiplatelet therapy in TRACER, we were lucky not to get to the disastrous oral GP IIb/IIIa inhibitor territory. The development of these agents has been associated not only with catastrophic bleeding disadvantage, but also more occlusive events including vascular deaths [10]. Importantly, in TRACER, allcause mortality and stent thrombosis trended against vorapaxar [4]. The FDA review revealed the unpublished evidence that vorapaxar in a daily dose of 2.5 mg causes N80% platelet inhibition in almost all patients by Day 7. However, the exposure-response and/or inhibition response relationships for efficacy and safety are unknown [9]. Assuming that these data are correct, it is impossible to comprehend why in the chronic setting we should cause platelet inhibition over 80% in non-STEMI patients by adding clopidogrel and aspirin on top of vorapaxar. The FDA also reports that bleeding time was measured in Phase 1 vorapaxar study with unvalidated assay, and at the time of the Agency review the impact of vorapaxar on bleeding time was unknown. We were not able to find any peer-reviewed published platelet study with vorapaxar either alone, or on top of aspirin and clopidogrel in patients after ACS, comparative data with prasugrel or ticagrelor, or at least retrieve some long-term data on platelet activity in healthy volunteers. It seems that TRACER was designed under pressure to start the trial as soon as possible without sufficient back-up of clinical and platelet data to justify such aggressive design. Moreover, overdosing prasugrel in TRITON repeatedly mentioned in the FDA reviews [6,7] and related publications (e.g. [11]) is mentally similar to applying triple antiplatelet therapy in TRACER suggesting a common mistake. In fairness, the TRACER leadership should be acknowledged for not repeating the TRITON pitfall by overdosing vorapaxar, preventing true disaster in TRACER. Finally, current restrictive indications for vorapaxar, and lack of approval in the USA for the NSTEMI patients are directly related to the pitfalls of TRACER design. It is entirely unclear whether or not applying dual antiplatelet strategies in TRACER will benefit vascular outcomes, however, TRITON data should be carefully considered for potential protocol adjustment(s). Vorapaxar per se is a mild platelet inhibitor with excellent safety profile in patients undergoing heart surgery, and renal failure [9], and it is unfortunate that the largest segment of patients
47
with coronary disease (NSTEMI) are not eligible, should now avoid, or use off-label Zontivity®. Disclosures Dr. Serebruany is listed as an inventor for the issued US patent “Treating vascular events with statins by inhibiting PAR-1 and PAR-4” (7,842,716) assigned to HeartDrug™ Research; and “Method for treatment of platelet activity with E5555” (USN 61/080,791); assigned to Eisai and HeartDrug™. He received compensation for the issued U.S. Patent 11/996,380 “Use of PAR-1/PAR-4 inhibitors for treating and preventing vascular diseases” on prasugrel assigned to Lilly. Dr Serebruany received funding for research studies with clopidogrel, and prasugrel; and consultant fees from clopidogrel manufacturers. Conflict of interest The authors report no relationships that could be construed as a conflict of interest. References [1] H.C. Diener, R. Sacco, S. Yusuf, Steering Committee, PRoFESSStudy Group, Rationale, design and baseline data of a randomized, double-blind, controlled trial comparing two antithrombotic regimens (a fixed-dose combination of extended-release dipyridamole plus ASA with clopidogrel) and telmisartan versus placebo in patients with strokes: the Prevention Regimen for Effectively Avoiding Second Strokes Trial (PRoFESS), Cerebrovasc. Dis. 23 (2007) 368–380. [2] H.C. Diener, J. Bogousslavsky, L.M. Brass, C. Cimminiello, L. Csiba, M. Kaste, D. Leys, J. Matias-Guiu, H.J. Rupprecht, MATCH investigators, Aspirin and clopidogrel compared with clopidogrel alone after recent ischaemic stroke or transient ischaemic attack in high-risk patients (MATCH): randomised, double-blind, placebo-controlled trial, Lancet 364 (2004) 331–337. [3] R.L. Sacco, H.C. Diener, S. Yusuf, D. Cotton, S. Ounpuu, W.A. Lawton, Y. Palesch, R.H. Martin, G.W. Albers, P. Bath, N. Bornstein, B.P. Chan, S.T. Chen, L. Cunha, B. Dahlöf, J. De Keyser, G.A. Donnan, C. Estol, P. Gorelick, V. Gu, K. Hermansson, L. Hilbrich, M. Kaste, C. Lu, T. Machnig, P. Pais, R. Roberts, V. Skvortsova, P. Teal, D. Toni, C. Vandermaelen, T. Voigt, M. Weber, B.W. Yoon, PRoFESS Study Group, Aspirin and extended-release dipyridamole versus clopidogrel for recurrent stroke, N. Engl. J. Med. 359 (2008) 1238–1251. [4] P. Tricoci, Z. Huang, C. Held, D.J. Moliterno, P.W. Armstrong, F. Van de Werf, H.D. White, P.E. Aylward, L. Wallentin, E. Chen, Y. Lokhnygina, J. Pei, S. Leonardi, T.L. Rorick, A.M. Kilian, L.H. Jennings, G. Ambrosio, C. Bode, A. Cequier, J.H. Cornel, R. Diaz, A. Erkan, K. Huber, M.P. Hudson, L. Jiang, J.W. Jukema, B.S. Lewis, A.M. Lincoff, G. Montalescot, J.C. Nicolau, H. Ogawa, M. Pfisterer, J.C. Prieto, W. Ruzyllo, P.R. Sinnaeve, R.F. Storey, M. Valgimigli, D.J. Whellan, P. Widimsky, J. Strony, R.A. Harrington, K.W. Mahaffey, the TRACER Investigators, Thrombin-receptor antagonist vorapaxar in acute coronary syndromes, N. Engl. J. Med. 366 (2012) 20–33. [5] S.D. Wiviott, E. Braunwald, C.H. McCabe, G. Montalescot, W. Ruzyllo, S. Gottlieb, F.J. Neumann, D. Ardissino, S. De Servi, S.A. Murphy, J. Riesmeyer, G. Weerakkody, C.M. Gibson, E.M. Antman, the TRITON-TIMI 38 Investigators, Prasugrel versus clopidogrel in patients with acute coronary syndromes, N. Engl. J. Med. 357 (2007) 2001–2015. [6] Prasugrel Secondary ReviewAvailable for download at http://www.fda.gov/ohrms/ dockets/ac/09/briefing/2009-4412b1-00-FDA.htm Assessed May 12, 2015. [7] Food and Drug Administration, Briefing Document, February 3, 2009 Meeting of FDA Cardiovascular and Renal Drugs Advisory Committee on PrasugrelAvailable for download at: http://www.fda.gov/downloads/advisorycommittees/committees meetingmaterials/drugs/cardiovascularandrenaldrugsadvisorycommittee/ucm181 185.pdf Assessed May 12, 2015. [8] G. Montalescot, S.D. Wiviott, E. Braunwald, S.A. Murphy, C.M. Gibson, C.H. McCabe, E.M. Antman, TRITON-TIMI 38 investigators, Prasugrel compared with clopidogrel in patients undergoing percutaneous coronary intervention for ST-elevation myocardial infarction (TRITON-TIMI 38): double-blind, randomised controlled trial, Lancet 373 (2009) 723–731. [9] NDA 294-886, Cross-discipline Team Leader Review on VorapaxarApril 18th, 2014 Available for download at: http://www.accessdata.fda.gov/drugsatfda_docs/nda/ 2014/204886Orig1s000SumR.pdf Assessed May 15, 2015. [10] D.P. Chew, D.L. Bhatt, S. Sapp, E.J. Topol, Increased mortality with oral platelet glycoprotein IIb/IIIa antagonists: a meta-analysis of phase III multicenter randomized trials, Circulation 103 (2001) 201–206. [11] V. Serebruany, D. Atar, Prasugrel: the real-life perspective, Circ. Cardiovasc. Qual. Outcomes 6 (2013) 253–254.
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
Editorial
Redesigning TRACER trial after TRITON Victor L. Serebruany ⁎ HeartDrug™ Research Laboratories, Johns Hopkins University, Osler Medical Building, 7600 Osler Drive, Suite 307, Towson, MD 21204, United States
a r t i c l e
i n f o
Article history: Received 16 May 2015 Received in revised form 13 June 2015 Accepted 16 June 2015 Available online 18 June 2015 Keywords: Clinical trials Trial design Vorapaxar Prasugrel Outcomes
a b s t r a c t Designing of smart clinical trials is critical for regulatory approval and future drug utilization. Importantly, trial design should be reconsidered if the interim analyses suggest unexpected harm, or conflicting results were yielded from the other trials within the same therapeutic area. With regard to antiplatelet agents, the perfect example is redesigning of the ongoing PRoFESS trial by eliminating aspirin from clopidogrel arm after the earlier MATCH trial results became available. The goal was to aseess the unchanged TRACER trial design in light of the evidence yielded from the earlier completed TRITON trial. TRACER was designed as a triple versus dual antiplatelet trial in NSTEMI patients with no previous long-term outcome data supporting such aggressive strategy. TRITON data represented dual versus dual antiplatelet therapy, and became available before TRACER enrollment starts revealing prasugrel front-loaded early vascular benefit predominantly in STEMI patients with the growing over time bleeding and cancer risks. Moreover, large prasugrel NSTEMI TRITON cohort exhibited trend towards excess mortality in experimental arm warning against aggressive TRACER design. The long-term TRITON results in general, and especially in the NSTEMI patients challenge unchanged TRACER trial design. Applying dual, rather than triple antiplatelet therapy protocol modification should be considered in TRACER to minimize bleeding, cancer, and non-cardiovascular death risks. © 2015 Published by Elsevier Ireland Ltd.
1. Introduction 1.1. Trial design Designing of clinical trials for regulatory drug assessment is a complex activity directly impacting chances for approval, future therapeutic options, and commercial success or failure. Importantly, trial designs are not written in stone, and are oftenly amended, adjusted, or even grossly changed. Usually, the main reasons leading to such dramatic trial redesign are unexpected findings discovered during interim analyses, or final results of other outcome-driven trials within the same therapeutic area. With regard to the antiplatelet agents, changes of then ongoing PRoFESS trial design are worth mentioning. Indeed, PRoFESS's initial objective sought to compare combination aspirin–dipyridamole to combination aspirin–clopidogrel [1]. However, after MATCH trial demonstrated that the addition of aspirin to clopidogrel had increased the risk of major bleeding complications without significantly reducing the risk of recurrent ischemic stroke [2], the PRoFESS protocol was subsequently modified. As a result, the 2027 patients initially randomized to receive aspirin–clopidogrel were taken off aspirin, and all the following patients randomized to that arm were administered clopidogrel monotherapy [3]. ⁎ Corresponding author at: 7600 Osler Drive, Suite 307 Yowson, MD 21204, USA. E-mail address: [email protected].
http://dx.doi.org/10.1016/j.ijcard.2015.06.020 0167-5273/© 2015 Published by Elsevier Ireland Ltd.
In short, trial design of ongoing active studies is a moving target, and it is the responsibility of trial leadership to watch closely for the constantly changing environment indirectly impacting expected clinical outcomes. Here, we will outline the comprehension of the Thrombin Receptor Antagonist for Clinical Event Reduction in Acute Coronary Syndrome (TRACER) trial [4] in light of the evidence yielded from the earlier TRial to assess Improvement in Therapeutic Outcomes by optimizing platelet InhibitioN with prasugrel Thrombolysis In Myocardial Infarction 38 (TRITON-TIMI-38) trial [5].
2. TRITON Landmark TRITON-TIMI 38 was a pivotal phase 3, randomized, double-blind, parallel-group, multinational, clinical study. The trial assessed head-to-head the efficacy and safety of the experimental antiplatelet agent prasugrel versus standard care with clopidogrel [5]. Patients (n = 13,608) with moderate to high-risk acute coronary syndromes (ACS) with the defined coronary anatomy undergoing percutaneous coronary intervention were randomized to prasugrel 60 mg loading dose followed by 10 mg daily or clopidogrel 300 mg loading dose followed by 75 mg daily for 6 to 15 months. The primary end point was the rate of cardiovascular (CV) death, myocardial infarction (MI), or stroke, and occurred in 12.1% of patients treated with clopidogrel, and 9.9% of patients randomized to prasugrel. This difference
Editorial
The trial randomized 12,944 patients with non-ST-elevation ACS in 37 countries throughout the world to receive vorapaxar (novel PAR-1 thrombin receptor antagonist) or placebo on top of standard care with clopidogrel and aspirin [4]. The symptom onset had to be within 24 h of presentation and eligible patients had to have either elevated troponins/creatinine kinase-myoglobin (CK-MB) or new ST-segment changes and at least one other high-risk marker such as prior MI or revascularization, peripheral vascular disease, or diabetes [4]. A composite of death from CV causes, MI, or stroke occurred in 822 patients in the vorapaxar group versus 910 in the placebo group (14.7% and 16.4%, respectively; hazard ratio, 0.89; 95% CI, 0.81 to 0.98; P = 0.02). Rates of moderate and severe bleeding were 7.2% in the vorapaxar group and 5.2% in the placebo group (hazard ratio, 1.35; 95% CI, 1.16 to 1.58;P b 0.001). Intracranial hemorrhage rates were 1.1% and 0.2%, respectively (hazard ratio, 3.39; 95% CI, 1.78 to 6.45; P b 0.001) in favor of placebo. Rates of nonhemorrhagic adverse events were similar in the two groups with the exception of excess of solid cancers after triple antiplatelet therapy [9]. Follow-up in the trial was terminated early after a safety review. 3.1. Chronology The time course of major activities for both trials is outlined in Table 1. The comparison of trial chronology suggests that the enrollment to TRACER starts after [4] the entire TRITON data became available [5]. Indeed, our comprehension of TRITON has been reevaluated after more balanced and detailed FDA reviews [6,7], than following primary overoptimistic publication [5]. However, all adverse but realistic TRITON findings were readily available before TRACER start, and could be carefully considered. 3.2. Target population Both trials share the dominant proportion of NSTEMI patients. In fact, over 94% in TRACER [4] and 74% in TRITON [5] experienced unstable angina, or non-STEMI ACS (NSTEMI) as qualifying event at enrollment. Since TRACER was designed specifically for NSTEMI patients, the outcome data, and especially TRITON mortality numbers in such target population are critically important, and should be considered before TRACER enrollment starts. Overall, in TRITON, while cardiovascular deaths slightly trended in favor of prasugrel (133 versus 150), however, excess in bleeding fatalities (21 versus 5), and 4 extra cancer-related deaths after prasugrel dilute the mortality benefit almost completely.
TRITON revealed few serious problems for prasugrel utilization as a chronic antiplatelet agent. In fact, there are 43 issues which should be considered for final TRACER design. First, despite the exclusion of heart surgery patients, those with high bleeding risks, and thrombocytopenia, prasugrel on top of aspirin cause highly significant increase of bleeding events than standard care with clopidogrel and aspirin combination [5–7]. In fact, the rates of key safety endpoint in TRITON, namely non-CABG major bleeding was 32% higher; spontaneous bleeds occur 52% more frequent; so as bleeding-related transfusions in the excess of 34% were observed in the prasugrel-treated patients. Moreover, fatal bleeding and major bleeding after CABG were over 4 times more often in the prasugrel arm. The second TRITON shortcoming was utilization of TIMI bleeding classification. This particular scale is the most conservative, focusing predominantly on major bleeding episodes, and excellently serving to assess safety of fibrinolytic agents. However, hemorrhagic strokes so prevalent with lytics are not common for modern antiplatelet strategies, when “insignificant” superficial bleeding episodes are most prevalent. By applying more liberal GUSTO bleeding classification, initial TRACER design will collect and report more bleeding events than filtered in TRITON. Since the TIMI scale-based existing TRITON bleeding rates after dual antiplatelet therapy were already known before TRACER start, applying triple antiplatelet therapy should be carefully comprehended for better and safer TRACER design. The third problem becomes obvious when considering the length of follow-up and bleeding risks in TRITON. Since the median follow-up in TRITON was 14.5 months, it was critical to realize when the most bleeding events occur. The FDA-generated Fig. 2 clearly answers this question with regard to the distribution of any bleeding events (A), and serious bleeding events (B). Indeed, there is an exponential growth of bleeding
Deaths in UA/NSTEMI Patients
0.00
Table 1 Chronology of TRITON and TRACER activities.
3.3. Bleeding
0.04
3. TRACER
Moreover, there were more deaths with unidentified causes after prasugrel (5 versus 2) in patients lost to follow-up without achieving primary endpoints, but affecting all-cause mortality [6,7]. In addition, the FDA provided a detailed distribution of deaths in TRITON, dependent on the type of qualifying initial occlusive event. The differences between mortality among patients with ST-elevated MI (STEMI), and UA/ NSTEMI in TRITON are presented in Fig. 1, and Table 1. There was a small but consistent trend (30 fatalities) towards more deaths in experimental TRITON arm among NSTEMI patients in contrast to STEMI cohort. This fact yielded from TRITON is a sufficient warning to be considered for TRACER design change.
fraction of patients 0.01 0.02 0.03
favoring experimental thienopyridine was predominantly attributed to the decrease of ischemic episodes during the first 3 days after intervention, and was especially evident in the STEMI cohort [6–8]. However, patients treated with prasugrel experienced higher bleeding rates, including significant increase of the key prespecified non-CABG-related TIMI major bleeding, and the reduced death from CV causes was canceled by the increased fatal bleeding on prasugrel [5–7], and surprising cancer signal after prasugrel [6,7].
45
Activity/trial
TRITON
TRACER
First patient in Last patient in Futility meeting MI definition changed Database locked Analyses completed Presentation/publication FDA reviews
November, 2004 January, 2007 July 2007 August 2007 September, 2007 October, 2007 November, 2007 February 2009
December, 2007 November, 2010 January, 2011 Not applicable September, 2011 October, 2011 November, 2011 April, 2014
0
2
4
6
Number at risk rx = Clopidogrel 5029 4867 4828 4735 rx = Prasugrel 5044 4896 4837 4742
8 month
10
12
14
4190 3562 3252 2767 4149 3542 3203 2695
rx = Clopidogrel
16 300 293
rx = Prasugrel
Fig. 1. Distribution of deaths over the follow-up in UA/NSTEMI patients in the TRITON trial. Exhibited from the FDA secondary review [6].
46
Editorial
Fig. 2. Distribution of any (A), and serious (B) bleeding events in TRITON. From the FDA reviews on prasugrel [7], and vorapaxar [9].
risks over time in the TRITON trial especially evident for serious bleedings. In fact, mean follow-up in TRACER was slightly longer (16 months) than in TRITON, increasing bleeding risks even further. Therefore, there is a compelling evidence that majority of bleeding events occur late, beyond the 30-days after initiation of therapy. This fact should have been kept in mind before finalizing TRACER protocol. Indeed, lack of direct comparisons, quoting bleeding incidence after prasugrel in TRITON does not negate the risk of bleeding with vorapaxar in TRACER. However, TRACER applied triple antiplatelet strategies, and if dual antiplatelet regiment in TRITON caused so much higher bleeding risks, then such obvious disadvantage should be kept in mind for potential modification of TRACER design after the TRITON data became available.
3.4. Cancer The excess of solid cancers after vorapaxar in TRACER (Hazard Ratio 1.4 (95% CI 1.1) top 1.8, p = 0.012) look very similar to those for prasugrel in TRITON [9] (see Fig. 3 for details). Importantly, the TRITON trial initiated the focus on the association of aggressive antiplatelet strategies, bleeding and cancer. In TRACER, as in TRITON, the recounted death numbers in all patients with solid cancer were high: 37 in vorapaxar patients vs. 24 in placebo [9]. As a rule of a thumb, more potent antiplatelet regimens cause more bleeding, and more
TRITON
solid cancers. Also, overall, the potency of antiplatelet therapy was high in TRACER, when triple antiplatelet therapy in the experimental arm (aspirin + clopidogrel + vorapaxar) exceeds 87% [4]. Finally, serious adverse events in TRACER were only to be reported until 60 days after the last dose of the study drug. While this limitation is less critical for adequate bleeding assessment, it is problematic for delayed event capturing, such as cancers, that take time to develop and be detected challenging triple antiplatelet strategies even further. Longer followup should be reinforced in future trials if we want to keep and assess delayed events such as cancers.
3.5. Impressions Indeed the trials have never been perfect, and every attempt should be made to identify the learning point from them in order to improve the next step. Obviously, there are limitations on extrapolation of the data from one trial to the other yet to start study. Indeed, any analogies may fail to work in practice and it is unclear that experience with a potent P2Y12 antagonist can be directly applied to a PAR-1 blocker. However, TRITON trial results in general, and especially in non-STEMI patients should warn against aggressive triple antiplatelet strategy later implemented in TRACER. The TRITON data were readily available before TRACER enrollment, but were not reconsidered resulting that the initial TRACER trial design has not been adjusted. The TRITON results
TRACER
Fig. 3. New solid cancers in TRITON and TRACER. From the FDA reviews on prasugrel [7], and vorapaxar [9].
Editorial
should impact TRACER protocol modification adding value to the increasing body of evidence suggesting that a more delicate strategy with modest platelet inhibition, especially in a chronic long-term setting, represents a substantial advantage over aggressive antiplatelet regimens. In this regard, applying triple antiplatelet strategies for the population with moderate ischemic burden (NSTEMI ACS), when excessive platelet inhibition is hardly justified was provocative, and contributes to the TRACER trial failure to gain NSTEMI indication. Ideally, TRACER should be designed with three arms: vorapaxar and aspirin; vorapaxar and clopidogrel; and versus standard care with clopidogrel and aspirin combination. If two arms will be the only practical option, then TRITON-type design directly comparing vorapaxar versus clopidogrel both on top of aspirin should apply. While distribution of thrombotic outcomes cannot be predicted, the bleeding and cancer signal will definitely look better for vorapaxar in TRACER. TRITON should not be overlooked for another important reason. Since clopidogrel substituted ticlopidine without any outcome-driven evidence, but rather on overcoming bone marrow toxicity safety concerns, TRITON was a first ever direct comparison of clinical outcomes between next generation thienopyridine and clopidogrel. If the TRACER leadership looks carefully at the TRITON results, they will realize that most if not all prasugrel benefit occur early, predominantly in STEMI cohort, with growing over time bleeding and cancer risks. The fact that all prasugrel patients at the end of TRITON were switched to clopidogrel reflects the stunning reality, and speaks for itself suggesting that TRITON leadership perfectly realize the escalating risks for keeping patients on prasugrel. Another issue is that when applying triple antiplatelet therapy in TRACER, we were lucky not to get to the disastrous oral GP IIb/IIIa inhibitor territory. The development of these agents has been associated not only with catastrophic bleeding disadvantage, but also more occlusive events including vascular deaths [10]. Importantly, in TRACER, allcause mortality and stent thrombosis trended against vorapaxar [4]. The FDA review revealed the unpublished evidence that vorapaxar in a daily dose of 2.5 mg causes N80% platelet inhibition in almost all patients by Day 7. However, the exposure-response and/or inhibition response relationships for efficacy and safety are unknown [9]. Assuming that these data are correct, it is impossible to comprehend why in the chronic setting we should cause platelet inhibition over 80% in non-STEMI patients by adding clopidogrel and aspirin on top of vorapaxar. The FDA also reports that bleeding time was measured in Phase 1 vorapaxar study with unvalidated assay, and at the time of the Agency review the impact of vorapaxar on bleeding time was unknown. We were not able to find any peer-reviewed published platelet study with vorapaxar either alone, or on top of aspirin and clopidogrel in patients after ACS, comparative data with prasugrel or ticagrelor, or at least retrieve some long-term data on platelet activity in healthy volunteers. It seems that TRACER was designed under pressure to start the trial as soon as possible without sufficient back-up of clinical and platelet data to justify such aggressive design. Moreover, overdosing prasugrel in TRITON repeatedly mentioned in the FDA reviews [6,7] and related publications (e.g. [11]) is mentally similar to applying triple antiplatelet therapy in TRACER suggesting a common mistake. In fairness, the TRACER leadership should be acknowledged for not repeating the TRITON pitfall by overdosing vorapaxar, preventing true disaster in TRACER. Finally, current restrictive indications for vorapaxar, and lack of approval in the USA for the NSTEMI patients are directly related to the pitfalls of TRACER design. It is entirely unclear whether or not applying dual antiplatelet strategies in TRACER will benefit vascular outcomes, however, TRITON data should be carefully considered for potential protocol adjustment(s). Vorapaxar per se is a mild platelet inhibitor with excellent safety profile in patients undergoing heart surgery, and renal failure [9], and it is unfortunate that the largest segment of patients
47
with coronary disease (NSTEMI) are not eligible, should now avoid, or use off-label Zontivity®. Disclosures Dr. Serebruany is listed as an inventor for the issued US patent “Treating vascular events with statins by inhibiting PAR-1 and PAR-4” (7,842,716) assigned to HeartDrug™ Research; and “Method for treatment of platelet activity with E5555” (USN 61/080,791); assigned to Eisai and HeartDrug™. He received compensation for the issued U.S. Patent 11/996,380 “Use of PAR-1/PAR-4 inhibitors for treating and preventing vascular diseases” on prasugrel assigned to Lilly. Dr Serebruany received funding for research studies with clopidogrel, and prasugrel; and consultant fees from clopidogrel manufacturers. Conflict of interest The authors report no relationships that could be construed as a conflict of interest. References [1] H.C. Diener, R. Sacco, S. Yusuf, Steering Committee, PRoFESSStudy Group, Rationale, design and baseline data of a randomized, double-blind, controlled trial comparing two antithrombotic regimens (a fixed-dose combination of extended-release dipyridamole plus ASA with clopidogrel) and telmisartan versus placebo in patients with strokes: the Prevention Regimen for Effectively Avoiding Second Strokes Trial (PRoFESS), Cerebrovasc. Dis. 23 (2007) 368–380. [2] H.C. Diener, J. Bogousslavsky, L.M. Brass, C. Cimminiello, L. Csiba, M. Kaste, D. Leys, J. Matias-Guiu, H.J. Rupprecht, MATCH investigators, Aspirin and clopidogrel compared with clopidogrel alone after recent ischaemic stroke or transient ischaemic attack in high-risk patients (MATCH): randomised, double-blind, placebo-controlled trial, Lancet 364 (2004) 331–337. [3] R.L. Sacco, H.C. Diener, S. Yusuf, D. Cotton, S. Ounpuu, W.A. Lawton, Y. Palesch, R.H. Martin, G.W. Albers, P. Bath, N. Bornstein, B.P. Chan, S.T. Chen, L. Cunha, B. Dahlöf, J. De Keyser, G.A. Donnan, C. Estol, P. Gorelick, V. Gu, K. Hermansson, L. Hilbrich, M. Kaste, C. Lu, T. Machnig, P. Pais, R. Roberts, V. Skvortsova, P. Teal, D. Toni, C. Vandermaelen, T. Voigt, M. Weber, B.W. Yoon, PRoFESS Study Group, Aspirin and extended-release dipyridamole versus clopidogrel for recurrent stroke, N. Engl. J. Med. 359 (2008) 1238–1251. [4] P. Tricoci, Z. Huang, C. Held, D.J. Moliterno, P.W. Armstrong, F. Van de Werf, H.D. White, P.E. Aylward, L. Wallentin, E. Chen, Y. Lokhnygina, J. Pei, S. Leonardi, T.L. Rorick, A.M. Kilian, L.H. Jennings, G. Ambrosio, C. Bode, A. Cequier, J.H. Cornel, R. Diaz, A. Erkan, K. Huber, M.P. Hudson, L. Jiang, J.W. Jukema, B.S. Lewis, A.M. Lincoff, G. Montalescot, J.C. Nicolau, H. Ogawa, M. Pfisterer, J.C. Prieto, W. Ruzyllo, P.R. Sinnaeve, R.F. Storey, M. Valgimigli, D.J. Whellan, P. Widimsky, J. Strony, R.A. Harrington, K.W. Mahaffey, the TRACER Investigators, Thrombin-receptor antagonist vorapaxar in acute coronary syndromes, N. Engl. J. Med. 366 (2012) 20–33. [5] S.D. Wiviott, E. Braunwald, C.H. McCabe, G. Montalescot, W. Ruzyllo, S. Gottlieb, F.J. Neumann, D. Ardissino, S. De Servi, S.A. Murphy, J. Riesmeyer, G. Weerakkody, C.M. Gibson, E.M. Antman, the TRITON-TIMI 38 Investigators, Prasugrel versus clopidogrel in patients with acute coronary syndromes, N. Engl. J. Med. 357 (2007) 2001–2015. [6] Prasugrel Secondary ReviewAvailable for download at http://www.fda.gov/ohrms/ dockets/ac/09/briefing/2009-4412b1-00-FDA.htm Assessed May 12, 2015. [7] Food and Drug Administration, Briefing Document, February 3, 2009 Meeting of FDA Cardiovascular and Renal Drugs Advisory Committee on PrasugrelAvailable for download at: http://www.fda.gov/downloads/advisorycommittees/committees meetingmaterials/drugs/cardiovascularandrenaldrugsadvisorycommittee/ucm181 185.pdf Assessed May 12, 2015. [8] G. Montalescot, S.D. Wiviott, E. Braunwald, S.A. Murphy, C.M. Gibson, C.H. McCabe, E.M. Antman, TRITON-TIMI 38 investigators, Prasugrel compared with clopidogrel in patients undergoing percutaneous coronary intervention for ST-elevation myocardial infarction (TRITON-TIMI 38): double-blind, randomised controlled trial, Lancet 373 (2009) 723–731. [9] NDA 294-886, Cross-discipline Team Leader Review on VorapaxarApril 18th, 2014 Available for download at: http://www.accessdata.fda.gov/drugsatfda_docs/nda/ 2014/204886Orig1s000SumR.pdf Assessed May 15, 2015. [10] D.P. Chew, D.L. Bhatt, S. Sapp, E.J. Topol, Increased mortality with oral platelet glycoprotein IIb/IIIa antagonists: a meta-analysis of phase III multicenter randomized trials, Circulation 103 (2001) 201–206. [11] V. Serebruany, D. Atar, Prasugrel: the real-life perspective, Circ. Cardiovasc. Qual. Outcomes 6 (2013) 253–254.