/    /  IV.5 Clinical Evaluations – Endpoints
Translation Pathway for Coronary Stent Development

Clinical Evaluations – Endpoints

Author

Donald E. Cutlip, MD

play-sharp-fill

Donald E. Cutlip, MD

Introduction

Unlike newer cardiac devices, coronary stents represent a mature technology, with over 20 years of randomized clinical trials assessing their safety and effectiveness. These clinical trials have resulted in the development and marketing of multiple iterations of bare-metal stents and several generations of drug-eluting stents (DES) as well as the recent disappointment of the first bioresorbable scaffold. In addition to specific technology advances, clinical trials of coronary stents have also tested adjunctive pharmacology, strategy comparisons versus medical therapy or coronary bypass surgery, and use in specific indications such as ST-segment elevation myocardial infarction (MI).

Along this journey, there have been great opportunities to develop and assess the clinical and nonclinical endpoints that provide the best measures of the safety and effectiveness of coronary stent technology. Knowledge gained from the results of these investigations has led to a continually evolving perspective on the true benefits and limitations of coronary stenting, which transcends device-related safety and effectiveness and allows improved decision making by physicians and patients for the most appropriate therapy in a variety of clinical scenarios. For example, in early stent trials, there was a focus on measures of restenosis as the most compelling measure of effectiveness and peri-procedural MI as the most relevant measure of safety. Although these endpoints remain of interest, subsequent studies have clarified that many events leading to death, MI, or revascularization were not related to the implanted stent or its patency, and that endpoints of coronary stent trials should also allow an assessment of overall patient outcomes in the setting of coronary artery disease treated with stents or other comparator therapies.

General Principles

Several general principles have emerged in the determination and definition of appropriate endpoints for coronary stent evaluations. For general acceptance, there must be consensus on meaningful endpoints among investigators, industry sponsors, and regulatory authorities, all of whom aim to incorporate what is important to patient care (1). The endpoints should be validated objective measures that can be assessed and compared using quantitative methods within a reasonably sized clinical trial. Ideally, the selection of endpoints will include those that meet regulatory requirements for device approval as well as for eventual reimbursement. Thus, the endpoints must allow for device-specific evaluation of performance related to safety and effectiveness as well as the positive or negative impact on global patient outcomes with some cost consideration or health care value assessment. Prior consensus efforts have also noted the importance of standardized endpoint definitions (1-3). Such standardization allows for formal pooled analyses or indirect comparison across devices and individual clinical trials, improves trial efficiency, and facilitates regulatory review. Finally, it is generally accepted that clinical endpoints within coronary stent trials should be adjudicated by an independent clinical event committee (4).

Role of Composites and Surrogates

Selection of individual endpoints to measure safety and effectiveness has represented a challenge in the design of efficient and reasonably sized clinical trials. Low frequency event rates or smaller differences between devices require larger clinical trials. Moreover, a selected individual endpoint such as target lesion revascularization (TLR) may fail to account for more serious competing risks of death and MI.

Composite endpoints have been used to overcome these limitations but must be used appropriately. There are three main concerns. First, the individual components of the composite should trend in the same direction. When this is the case, the increased composite event rate should increase statistical power and reduce the sample size required. When this is not the case, use of composites may actually obscure a specific benefit or harm and reduce statistical power. Second, the reduced sample size prevents an adequately powered assessment of individual components of the composites, which at times may be important for overall safety or effectiveness analysis. Third, it is ideal if the different weights of the components can be assigned based on severity and importance to patients. This has usually not been the case in coronary stent trials, and the use of composite endpoints has led to trial results being driven by an endpoint of lower clinical importance, but wrongly interpreted as also showing a difference in more severe components as well.

Surrogate endpoints have also been considered as a means for designing smaller and more efficient trials but have fallen short of the needs for regulatory approval. For example, late lumen loss based on quantitative coronary angiography represents a reasonable surrogate for clinical restenosis (5), but it provides insufficient safety data to be considered as a primary endpoint in pivotal clinical trials. More recently, there has been interest in use of intravascular ultrasound or optical coherence tomography to obtain surrogate measures of stent outcomes (6,7). These imaging modalities offer promise for assessing optimal stent deployment and measures of vessel healing but lack the rigor needed to support safety or clinical effectiveness.

Assessing Value and Patient-reported Outcomes

Coronary stent trials generally have been designed to meet the requirements for regulatory market approval, and endpoints have been selected to provide evidence for a reasonable assurance of safety and effectiveness.  Because of this objective, the endpoints have been centered on device performance. More recently, there has been increased interest in the assessment of coronary stents, as well as other devices, for their impact on global patient outcomes and overall health care value. In the case of coronary stents, previous analyses of patient outcomes from a health care value perspective have utilized traditional cost-effectiveness methods with an output of cost per quality-adjusted life-year (QALY).  While these analyses have supported reimbursement decisions according to a metric of reasonable and necessary, both the quality and cost sides of the equation have been based mostly on device-oriented endpoints.

Patient-oriented outcomes such as all-cause mortality or stroke are clearly of interest to patients, and devices showing benefit in these outcomes even at high costs would likely be considered reasonable and necessary. However, differences between therapies for these endpoints are seldom present in contemporary studies, and, if present, are difficult to measure in a reasonably sized trial. Patient-reported outcome measures (PROMs) provide a method to assess other events that may have less dramatic impact on vital status but have a meaningful impact on quality of life.  Validated patient-reported outcomes assessment instruments not only provide an additional evaluation of events that are important to patients, but also allow discrimination between therapies that may appear similar based on traditional clinical endpoints.  Although it is unlikely PROMs will be adequate as primary endpoints for regulatory approval, they are potentially important as components of composite endpoints with hierarchical weighting or as standalone secondary endpoints. In 2009, FDA issued a guidance document on the appropriate use of PROMs (8), while a more recent guidance document provided the Agency’s current thinking on the use of patient preferences in regulatory decision-making (9). As regulatory approval authorities and reimbursement agencies encourage parallel review of study data targeting device approval and reimbursement, it is likely PROMs will be increasingly incorporated into clinical trial designs for value-based hypothesis testing.

It is necessary that PROMs as potential clinical trial endpoints meet standards of validity (evidence for measurement of the concept of interest), reliability (results are reproducible), and discrimination (detect change and meaningful differences between groups). The United States Food and Drug Administration (FDA) has qualified the Kansas City Cardiomyopathy Questionnaire (KCCQ) and the Minnesota Living with Heart Failure Questionnaire (MLHFQ) for potential use by medical device sponsors in the development and evaluation of medical devices (10).

Although patient-centered outcomes are key to assessing health care value, refinement of efforts for measuring the cost impact of new technology is also needed. Using traditional cost-effectiveness analyses reported in QALYs, it has been suggested that a cost <$50,000 per QALY would represent high value while a cost >$150,000 per QALY would be low value (11). While these analyses are useful for health care decision making, value-based sustainability also requires consideration of overall costs related to anticipated market demands.

Impact of Trial Designs on Selection and Assessment of Endpoints

Most historical information on clinical endpoints related to coronary stents has come from fairly simple trials designed to compare various stent types in selected patient populations. There are now increased numbers of trials designed to compare stents to alternative medical or surgical therapies or to compare adjunctive pharmacological therapies among patients receiving coronary stents. Endpoints for these trials must allow for measuring meaningful differences given the relevant comparators. In some cases, this may mean defining criteria for endpoints that are consistent with the therapies being tested (e.g., defining procedure-related MI in a stent versus coronary artery bypass surgery trial) or selecting primary endpoints that are more likely to measure differences between groups, such as stroke or bleeding endpoints in a trial comparing antithrombotic agents.

Evolution in coronary stent development has necessitated several other modifications in clinical trial design that may impact endpoint selection or assessment. Given progressively lower event rates and narrowing rate differences between devices, it has become increasingly difficult to demonstrate clinical or statistical superiority or noninferiority among study devices in reasonably sized studies; thus, larger sample sizes are required. It is also possible that differences in some endpoints may emerge over longer periods of follow-up.  For example, despite low target lesion failure (TLF) (a composite of cardiac death, target vessel MI, and TLR) rates for contemporary DES at 1 year, there is a persistent risk of 3% per year during follow-up (12). A new device that improves this late outcome would be beneficial but would require long-term follow-up to demonstrate a difference. Larger trials with longer-term follow-up will require simplification in data collection and monitoring to avoid excessive costs, and this may impact reporting and verification of some endpoints.  In some cases, by appropriately balancing pre- and post-market data collection, key effectiveness endpoints may be assessed in smaller premarket studies, while the assessment of low frequency and/or late safety event rates may be completed in the post-market setting (13).

Specific Endpoints

Device-oriented Endpoints

Safety and effectiveness endpoints that are related to the device remain integral to the assessment of safety and effectiveness. Table 1 shows currently recommended endpoints. FDA has recommended using TLF as the primary endpoint for most stent versus stent studies. In addition, it is important with any composite to report the rate of the individual components of the composite to demonstrate consistency among the components and to avoid inappropriate claims of a difference in more serious outcomes such as death or MI when TLR is the driving factor. The use of the composite offers advantages of an increased number of events and accounts in part for competing risks of death and MI when reporting outcomes. For defining cardiac-related death, we recommend the definition described by the FDA’s Standardized Data Collection Initiative (3). This includes death due to all cardiovascular causes such as MI, heart failure, arrhythmia, or sudden deaths as well as vascular causes of pulmonary embolism or stroke. In some cases, investigators may wish to separate definite cardiac and vascular causes.

 

Table 1. Recommended Device-oriented Endpoints and Definitions

Endpoint/Composite Recommended Standard Definition Adjudication Comments
Target lesion failure (TLF) Cardiac death, target vessel MI, TLR Target vessel MI unless electrocardiogram or imaging evidence clearly indicates non-target vessel
Target vessel failure (TVF) Cardiac death, target vessel MI, TVR
Cardiac death Death resulting from an acute MI, sudden cardiac death, death due to heart failure, death due to stroke, death due to cardiovascular (CV) procedures, death due to CV hemorrhage, and death due to other CV causes. Death is classified as cardiac based on initiating event and includes all deaths related to cardiac procedure or complication thereof.
Myocardial infarction (MI) Should be used when there is evidence of myocardial necrosis in a clinical setting consistent with myocardial ischemia. Requires stable or falling baseline value for biomarker.
Target lesion revascularization (TLR) Clinically driven* or ischemia driven† Diameter stenosis should be based on angiographic core laboratory assessment if available.
Target vessel revascularization (TVR) Clinically driven* or ischemia driven†
Stent (scaffold) thrombosis Academic Research Consortium Definite or Probable Generally requires review of angiogram or angiographic core lab assessment.

*Clinically driven includes ischemic symptoms and either diameter stenosis >50%, fractional flow reserve (FFR) <0.80, or instantaneous wave-free ratio (iFR) <0.89; for studies with routine angiographic follow-up we include core lab diameter stenosis >70% even in the absence of symptoms or other functional data.

†Ischemia driven includes FFR <0.80 or iFR <0.89 or positive noninvasive test for ischemia.

There remains controversy over the definition of periprocedural MI for coronary stent trials, but we recommend the definition for clinically relevant MI proposed by Society of Cardiac Angiography and Intervention of CK-MB more than 10 times upper limit of normal (or 5 times upper limit of normal in the presence of new Q waves); if CK-MB is not available then troponin more than 70 or 35 times upper limit of normal, respectively, is recommended (14). For spontaneous MI >48 h after the procedure and for other events not related to percutaneous coronary intervention or coronary artery bypass surgery, we recommend the definitions proposed by the Joint Task Force for Universal Definition of Myocardial Infarction and subsequent updates (15). For TLR endpoints, we recommend reporting clinically- or ischemia-driven events. Stent or scaffold thrombosis should be defined as probable or definite by Academic Research Consortium (ARC) criteria. For endpoints defined according to the treated lesion, we recommend inclusion of the original stented segment plus the 5 mm immediately distal or proximal to the original stented segment. For endpoints defined according to the target vessel, we recommend this include the entire epicardial vessel in which the stent was placed as well as its side branches. In the case of left main coronary artery treatment, the target vessel includes the entirety of the left anterior descending and left circumflex coronary arteries.

Patient-oriented Endpoints

Patient-oriented endpoints are critical for assessing safety outcomes that may be unrelated to specific device performance. Ideally, they should assess the net positive or negative impact of the coronary stent procedure, given the presence of underlying coronary artery disease and other comorbidities. Recommended patient-oriented composite and individual endpoints are shown in Table 2. The composite of all-cause mortality, any MI, and any repeat revascularization has been recommended by ARC as a broad measure of clinical outcome that is likely to be related to complications of coronary disease and its treatment.

 

Table 2. Patient-oriented Endpoints and Recommended Definitions

Endpoint/Composite Recommended Standard Definition Adjudication Comments
Patient-oriented composite All-cause death, any myocardial infarction (MI), or any revascularization MI adjudication criteria as for device-oriented endpoint; Revascularization included regardless of indication.
Bleeding Bleeding ARC (BARC) class 2-5 Adjudication should specify BARC class and any other protocol-based bleeding definitions.
Stroke Acute episode of focal or global neurological dysfunction caused by brain, spinal cord, or retinal vascular injury as a result of hemorrhage or infarction. Adjudication committee should include a stroke neurologist.
Patient-reported outcomes (PRO) Measurements based on a report that comes directly from the patient (i.e., study subject) about the status of a patient’s health condition without amendment or interpretation of the patient’s response by a clinician or anyone else.  A PRO can be measured by self-report or by interview provided that the interviewer records only the patient’s response. Scores should be validated by a core laboratory experienced in the use of the specific tool.

Stroke and bleeding are important patient-centered endpoints but have been given less attention in coronary stent trials. Fortunately, stroke is infrequent, but given the disastrous impact on quality of life it should be clearly defined and reported. Recently, there have been efforts to standardize stroke definitions in cardiac device trials (3,16,17), and we recommend using a classification based on these efforts from the FDA’s Standardized Data Collection Initiative and ARC. Stroke is defined as an acute episode of focal or global neurological dysfunction caused by brain, spinal cord, or retinal vascular injury as a result of hemorrhage or infarction. Stroke may be distinguished from transient ischemic attacks based on persistence of symptoms or evidence of infarction based on imaging. We recommend an event duration of >24 h or death before 24 h to classify as a stroke and include events lasting <24 h if confirmed by imaging. All events should be adjudicated by a qualified stroke neurologist. It may also be useful to classify strokes as ischemic or hemorrhagic. Ischemic strokes that undergo hemorrhagic transformation should be classified as ischemic or within a specific category noting hemorrhagic transformation depending on the study. Subdural hematoma or intracranial hemorrhagic events are not strokes but are important to capture, since coronary stent patients routinely take anti-thrombotic therapy. In some studies, it may also be useful to classify strokes according to severity of disability. We recommend using the modified Rankin scale assessed at 90 days for this determination.

Bleeding after coronary stenting has been recognized to have serious consequences with an impact on subsequent mortality as least as great as spontaneous MI (18). Assessing the frequency and associated outcomes of bleeding has been difficult due to a lack of standardized reporting. We recommend use of the classification proposed by the Bleeding ARC (BARC) and suggest reporting BARC class 2-5 as clinically significant bleeding and class 3-5 as major bleeding (19).

Finally, we recommend the use of PROMs as important secondary patient-oriented endpoints. Although clinical outcomes remain paramount in assessing the safety and effectiveness of coronary stents and other cardiac devices, differences in validated measures of symptoms or quality of life may prove critical in discriminating between devices that appear similar based on traditional clinical events.

References

  1. Cutlip DE, Windecker S, Mehran R, et al. Clinical end points in coronary stent trials: a case for standardized definitions. 2007;115:2344-51.
  2. Hicks KA, Tcheng JE, Bozkurt B, et al. 2014 ACC/AHA Key Data Elements and Definitions for Cardiovascular Endpoint Events in Clinical Trials: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Data Standards (Writing Committee to Develop Cardiovascular Endpoints Data Standards). J Am Coll Cardiol. 2015;66:403-69.
  3. Hicks KA, Mahaffey KW, Mehran R, et al. 2017 Cardiovascular and Stroke Endpoint Definitions for Clinical Trials. 2018;137:961-72.
  4. Seltzer JH, Heise T, Carson P, et al. Use of endpoint adjudication to improve the quality and validity of endpoint assessment for medical device development and post marketing evaluation: Rationale and best practices. A report from the cardiac safety research consortium. Am Heart J. 2017;190:76-85.
  5. Mauri L, Orav EJ, Kuntz RE. Late loss in lumen diameter and binary restenosis for drug-eluting stent comparison. 2005;111:3435-42.
  6. Brown BG, Zhao XQ. Is intravascular ultrasound the gold standard surrogate for clinically relevant atherosclerosis progression? J Am Coll Cardiol. 2007;49:933-8.
  7. Tahara S, Chamié D, Baibars M, Alraies C, Costa M. Optical coherence tomography endpoints in stent clinical investigations: strut coverage. Int J Cardiovasc Imaging. 2011;27:271-87.
  8. S. Food and Drug Administration. Patient-Reported Outcome Measures: Use in Medical Product Development to Support Labeling Claims. December 2009. Available at: https://www.fda.gov/media/77832/download. Accessed May 4, 2019.
  9. S. Food and Drug Administration. Patient Preference Information – Voluntary Submission, Review in Premarket Approval Applications, Humanitarian Device Exemption Applications, and De Novo Requests, and Inclusion in Decision Summaries and Device Labeling. October 2016. Available at: https://www.fda.gov/media/92593/download. Accessed May 4, 2019.
  10. S. Food and Drug Administration. Medical Device Development Tools (MDDT). Last updated March 12, 2019. Available at: https://www.fda.gov/MedicalDevices/ScienceandResearch/MedicalDeviceDevelopmentToolsMDDT/default.htm. Accessed May 4, 2019.
  11. Anderson JL, Heidenreich PA, Barnett PG, et al. ACC/AHA statement on cost/value methodology in clinical practice guidelines and performance measures: a report of the American College of Cardiology/American Heart Association Task Force on Performance Measures and Task Force on Practice Guidelines. J Am Coll Cardiol. 2014;63:2304-22.
  12. Vlachojannis GJ, Smits PC, Hofma SH, et al. Long-term clinical outcomes of biodegradable polymer biolimus-eluting stents versus durable polymer everolimus-eluting stents in patients with coronary artery disease: three-year follow-up of the COMPARE II (Abluminal biodegradable polymer biolimus-eluting stent versus durable polymer everolimus-eluting stent) trial. 2015;11:272-9.
  13. S. Food and Drug Administration. Balancing Premarket and Postmarket Data Collection for Devices Subject to Premarket Approval. April 2015. Available at: https://www.fda.gov/media/88381/download. Accessed May 4, 2019.
  14. Moussa ID, Klein LW, Shah B, et al. Consideration of a new definition of clinically relevant myocardial infarction after coronary revascularization: an expert consensus document from the Society for Cardiovascular Angiography and Interventions (SCAI). Catheter Cardiovasc Interv. 2014;83:27-36.
  15. Thygesen K, Alpert JS, Jaffe AS, et al. Third universal definition of myocardial infarction. J Am Coll Cardiol. 2012;60:1581-98.
  16. Lansky AJ, Messé SR, Brickman AM, et al. Proposed Standardized Neurological Endpoints for Cardiovascular Clinical Trials: An Academic Research Consortium Initiative. J Am Coll Cardiol. 2017;69:679-91.
  17. Kappetein AP, Head SJ, Généreux P, et al. Updated standardized endpoint definitions for transcatheter aortic valve implantation: the Valve Academic Research Consortium-2 consensus document. J Thorac Cardiovasc Surg. 2013;145:6-23.
  18. Brener SJ, Kirtane AJ, Stuckey TD, et al. The Impact of Timing of Ischemic and Hemorrhagic Events on Mortality After Percutaneous Coronary Intervention: The ADAPT-DES Study. JACC Cardiovasc Interv. 2016;9:1450-7.
  19. Mehran R, Rao SV, Bhatt DL, et al. Standardized bleeding definitions for cardiovascular clinical trials: a consensus report from the Bleeding Academic Research Consortium. 2011;123:2736-47.
Hide picture