Author + information
- Karen A. Hicks, MD∗ ( and )
- Robert J. Temple, MD
- Office of Drug Evaluation 1, Center for Drug Evaluation and Research, U.S. Food and Drug Administration, Silver Spring, Maryland
- ↵∗Address for correspondence:
Dr. Karen A. Hicks, U.S. Food and Drug Administration, Division of Cardiovascular and Renal Products, Office of Drug Evaluation 1, Center for Drug Evaluation and Research, 10903 New Hampshire Avenue, Building 22, Room 4182, Silver Spring, Maryland 20993-0002.
- myocardial infarction after coronary revascularization
- myocardial infarction related to PCI
- periprocedural myocardial infarction
- type 4a myocardial infarction
Myocardial infarction (MI) is a commonly used endpoint in clinical trials evaluating the efficacy and safety of cardiovascular drugs and devices. Periprocedural MIs (PMIs), also known as type 4a MIs, contribute to the MI endpoint and can account for a substantial number of events. In the TRITON–TIMI 38 (Trial to Assess Improvement in Therapeutic Outcomes by Optimizing Platelet Inhibition With Prasugrel–Thrombolysis In Myocardial Infarction 38) study, for example, approximately 40% of MIs occurred periprocedurally (1). How a PMI is defined influences the number of MI events that may be adjudicated by a clinical events committee (CEC).
In this issue of JACC: Cardiovascular Interventions, Spitzer et al. (2) develop algorithms to identify PMI events in the RAC (RESOLUTE All-Comers) trial using 3 definitions currently used in percutaneous coronary intervention trials and demonstrate that PMI definitions and the criteria used to identify PMI events (i.e., event triggers) are not interchangeable.
The RAC trial was a prospective, multicenter, randomized, noninferiority trial comparing the Medtronic Resolute Zotarolimus-Eluting Coronary Stent System (Medtronic, Minneapolis, Minnesota) with the Abbott XIENCE V Everolimus-Eluting Coronary Stent System (Abbott Vascular, Santa Clara, California) in 2,292 subjects at 17 Western European sites. The primary endpoint was target lesion failure, defined as a composite of cardiac death, MI not clearly attributable to a nontarget vessel, and clinically indicated target lesion revascularization at 12 months (3). Follow-up occurred out to 5 years. According to the instructions for use, “data from the RAC trial were used to support the pre-market approval (PMA) of the Resolute Integrity stent” (4).
The 3 definitions used for the PMI algorithms were: 1) the World Health Organization extended definition (WHO-ED); 2) the 2012 third universal definition of MI (TUD) of the Joint European Society of Cardiology, American College of Cardiology, American Heart Association, and World Health Federation Task Force for the Universal Definition of Myocardial Infarction; and 3) the Society for Cardiovascular Angiography and Interventions (SCAI) definition of clinically relevant MI after coronary revascularization (5–7). Table 1 summarizes these definitions and their criteria. The WHO-ED was developed out of necessity during CEC adjudication of the RAC trial when, after 70% of the patients were recruited, it “became evident that CK [creatine kinase] and CK-MB, the study criteria for MI,” were not available “in a sizeable proportion of patients because of biomarker collection compliance issues” (8). The WHO-ED was thus created to allow PMI adjudication even when biomarker data are incomplete and to allow a bridge to the historical control data from the Medtronic ENDEAVOR clinical program (which used CK and CK-MB) (5).
As Spitzer et al. (2) note, these definitions are fundamentally different, with distinct preferred biomarkers (CK, CK-MB, cardiac troponin), thresholds (>2, >3, >5, and >10 times the upper limit of normal), and different clinical, electrocardiographic (ECG), imaging, and angiographic criteria required to define PMI. Given these differences, it is not surprising that the number of PMI events for any given dataset varies with these definitions.
The investigators identify 5 triggers (i.e., on the basis of investigator reports, monitor findings, electronic case report forms, ECG and angiographic core laboratory findings, and CEC reviews) to detect potential PMI events. These triggers generally led to requests for additional source documents to create adjudication packets for the CEC to review.
The investigators also highlight the following 6 specific issues considered in developing the PMI algorithms: 1) the interchangeability of upper reference limit and upper limit of normal units; 2) whether the biomarkers were stable, rising, or falling and collection of biomarkers was incomplete; 3) presence of MI at baseline; 4) use of ECG data specifically for the WHO-ED and the SCAI definition and presuming that ECG changes were present when tracings were missing; 5) the hierarchical use of biomarkers; and 6) the presence or absence of MI at baseline when no cardiac biomarkers were available post-procedure. The resulting algorithms appear to represent the selected definitions adequately.
Of the 2,509 procedures included in this analysis, 382, 234, and 216 were associated with PMI triggers according to the TUD, the WHO-ED, and the SCAI definition, respectively. A total of 636 PMI triggers were identified, only 38 (6%) of which were common to all definitions; 478 unique triggers were also identified, including 240 (37.7%), 91 (14.3%), and 147 (23.1%) using the TUD, the WHO-ED, and the SCAI definition, respectively. Concordance between 2 definitions was highest between the TUD and WHO-ED (14.0%) but was similar between the TUD and the SCAI definition (2.4%) and between the WHO-ED and the SCAI definition (2.5%).
More importantly, additional analyses demonstrated that when clinical presentation and ECG data were unavailable, the number of triggers increased by 40% for all definitions combined, because these data caused some elevated biomarkers to no longer support PMI. Hence, missing data increase the number of triggers and potential PMI events. Similarly, an analysis using only the peak post-procedural value of a biomarker (instead of all post-procedural values) increased the number of triggers from 636 to 867, except for the SCAI definition, for which the number of triggers was similar (214 with peak value vs. 216 with all available values). In summary, when clinical presentation, ECG data, and all post-procedural cardiac biomarker data are available, the number of triggers for PMI decreases with all 3 definitions.
How the data are collected throughout a clinical trial is critical. Although the investigators suggest that “the SCAI definition may be preferred for large [randomized controlled trials] due to its simplicity and less costs involved, as well as in strategy trials comparing [percutaneous coronary intervention] versus [coronary artery bypass grafting],” we recommend a different approach. The U.S. Food and Drug Administration’s Center for Drug Evaluation and Research generally uses the TUD but recognizes that collecting data to facilitate analyses (preferably pre-specified) with alternative MI definitions (such as that of SCAI) may be reasonable and can confirm consistency of signals.
In summary, MI is a commonly used endpoint in clinical trials. PMIs contribute to the MI endpoint and in some cases can account for a substantial fraction of the events. How a PMI is defined can influence the number of MI events that are adjudicated by the CEC. Although Spitzer et al. (2) demonstrate that the TUD, the WHO-ED, and the SCAI PMI definition and their event triggers are not interchangeable, more importantly, they demonstrate that missing data increase the number of triggers and potential number of PMI events for each of the definitions. We encourage early (before study initiation) interaction with the U.S. Food and Drug Administration to discuss the proposed trial and supporting documentation, because the devil is always in the details. We also recommend using the 2014 American College of Cardiology and American Heart Association’s “Key Data Elements and Definitions for Cardiovascular Endpoint Events in Clinical Trials” for regulatory submissions that are based on work by the Standardized Data Collection for Cardiovascular Trials Initiative (10).
Rigorous collection of clinical, ECG, angiographic, imaging, and laboratory results may improve the correct identification of PMI.
↵∗ Editorials published in JACC: Cardiovascular Interventions reflect the views of the authors and do not necessarily represent the views of JACC: Cardiovascular Interventions or the American College of Cardiology.
This editorial reflects the views of the authors and should not be construed to represent the views or policies of the U.S. Food and Drug Administration. Both authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- ↵U.S. Food and Drug Administration. Prescribing information for Effient. Available at: http://www.accessdata.fda.gov/drugsatfda_docs/label/2011/022307s003lbl.pdf. Accessed March 1, 2017.
- Spitzer E.,
- de Vries T.,
- Cavalcante R.,
- et al.
- ↵Medtronic. Instructions for use (Resolute Integrity Zotarolimus-Eluting Coronary Stent System). Minneapolis, MN: Medtronic, 2010.
- Thygesen K.,
- Alpert J.S.,
- Jaffe A.S.,
- Simoons M.L.,
- Chaitman B.R.,
- White H.D.,
- on behalf of the Joint ESC/ACCF/AHA/WHF Task Force for the Universal Definition of Myocardial Infarction
- Moussa I.D.,
- Klein L.W.,
- Shah B.,
- et al.
- Cavalcante R.,
- Serruys P.W.
- White H.
- Hicks K.A.,
- Tcheng J.E.,
- Bozkurt B.,
- et al.