PART IV - CONSENSUS ON DEFINITIONS OF CLINICAL ENDPOINTS: PERCUTANEOUS CORONARY AND VALVULAR INTERVENTION TRIALS
TABLE OF CONTENTS
BOOKSHOP
Updated on May 14, 2021
PART IV

Consensus on definitions of clinical endpoints: percutaneous coronary and valvular intervention trials

Pascal Vranckx1, Donald E. Cutlip, Roxana Mehran2, Martin B. Leon3, Patrick W. Serruys4
1 Hartcentrum Hasselt - Hasselt, Belgium
2Icahn School of Medicine at Mount Sinai - New York, United States
3Columbia University Medical Center - New York, United States
4Department of Cardiology, National University of Ireland, Galway (NUIG) - Galway, Ireland

Summary

Controlled clinical trials are a cornerstone in the search for medical therapeutic advances. The process by which clinical trials in cardiovascular medicine are designed, conducted, analysed, presented, and published has evolved dramatically over the last decade.

In order to evaluate the effects of a particular treatment strategy on mortality and major morbidity within a disease entity, large, global studies with relatively long-term clinical endpoints are conducted. One of the substantial challenges in the conduct of clinical trials relates to the definition, collection, and accurate assessment of endpoint data in a consistent and timely manner.

Globalisation in percutaneous cardiovascular intervention research calls for standardisation and uniform definition of clinical endpoints and the harmonisation of clinical event adjudication. Consensus on uniform definitions is required to enable meaningful, albeit informal and indirect, comparisons between clinical trials and to aggregate outcome data within trial programmes and beyond. This process must be based on considerations ranging from historical legacy to key pathophysiologic mechanisms and relevance to clinical interpretability.

The purpose of this chapter is to introduce the interested reader to the process of consensus clinical endpoint definitions generation, to provide an appraisal of the most commonly used endpoints and reference to the context of their rationale. We will reflect on their impact on actual trial design, conduct and reporting.

Introduction

Controlled safety and efficacy endpoint trials are a cornerstone in the search for medical therapeutic advances, their approval by regulatory bodies, and their adoption for clinical use by the medical communities driven by clinical practice guidelines. The concern about how the effects of therapies is understood in clinical practice requires attention to the methods of randomised clinical trials, including the selection of meaningful clinical endpoints.

Over the past decade, substantially more clinical trials rely on the adjudication of safety and efficacy endpoints by independent clinical event committees (CEC, also known as clinical endpoint committees). The mission of the CEC includes ensuring the quality and integrity of the trial by maintaining uniformity, accuracy, and consistency in the interpretation of clinical endpoints across geographical areas and over the long course of the study. This assumes the implementation of a readily assessable, validated, and preferably blinded process for event definition that provides data sufficiently specific to be subjected to statistical analysis.

Inevitably, consensus on endpoint definitions will include certain arbitrary features. For this and other reasons, they should not be considered practice standards or be confused with guidelines. Rather, their intent should be to provide an instrument for standardisation of clinical trial methods and allow for indirect trial-to-trial comparisons as well as to facilitate the pooling of data involving multiple investigations.

FOCUS BOX 1The need for a consensus on endpoint definitions
  • Uniform, accurate and consistent event adjudication requires independent event adjudication based on trial specific or consensus definitions available in the field, providing outcome data sufficiently specific to be subjected to statistical analysis
  • One of the most fundamental barriers to meaningful data aggregation is the use of different nomenclature and definitions of key descriptors and endpoints from one trial to another

Consensus on definitions in percutaneous cardiovascular interventional trials: The consensus methodology

In many ways, the interventional cardiovascular community has created a model for translational medicine by incorporating industry, academia, regulatory bodies [the United States’ Food and Drug Administration (FDA)/ The European Medicines Agency (EMEA), and clinical practitioners into an on-going effort to standardise and define endpoints used in randomised clinical trials.

This process is exemplified by the founding in 2005 of the Academic Research Consortium (ARC), an informal collaboration of academic research organisations, clinicians, device manufacturers, and the FDA that introduced uniform definitions for drug-eluting stent clinical trials [1, 2]. This process has been utilised to provide consensus definitions for bleeding (BARC), low extremity peripheral artery disease (PARC), medication non adherence (NARC), neurologic events (NeuroARC), more recently High Bleeding Risk (HBR), and clinical trials of percutaneous aortic valves[3, 4, 5, 6, 7, 8, 9, 10, 11].

In this paragraph, we will highlight the Valve Academic Research Consortium (VARC) model as an example of this process, as it involved even more complex relationships between surgical and medical specialties. The VARC was established in 2009 to propose standardised consensus clinical endpoints and definitions for use in transcatheter aortic valve implantation (TAVI) investigations. The writing committee consisted of acknowledged experts in the field of interest, as well as liaisons from the regulatory bodies, the peer specialist societies (cardiology and cardiac surgery) in the USA and Europe and academic research organisations with a common interest and established expertise in cardiovascular trials [2]. Over a series of meetings and conference calls, the writing committee discussed the topics to be covered in the document and assigned lead authors for each section. Section outlines of the master document were drafted integrating evidence obtained by literature searches and/or recommendations of dedicated working groups within professional societies. Of note, teleconferences were scheduled between the writing committee chair and members who were not present at the meetings to ensure consensus on the document. After the writing committee had reviewed each section, discussed overall document content, and ultimately arrived at a consensus, the master document was sent for external peer review. Following peer review, the writing committee chair engaged authors to address reviewer comments and finalise the document for notification and/or organisational approval. The VARC made every effort to avoid any actual or potential conflicts of interest that might arise as a result of an outside relationship or personal interest of a member of the writing panel. Since its first publication a VARC definition update and a MVARC (mitral valve research) were issued [4, 5] implementing a similar format.

Criteria for endpoint definitions: The VARC model

Introduction of new cardiovascular devices should ideally follow a standard bench-to-bedside evidence-based medicine pathway, starting with preclinical testing and advancing to meaningful clinical investigations. Endpoint definitions in percutaneous cardiovascular intervention trials need to be characterised to the specific implant, the implant procedure, and the resultant patient oriented outcomes, which can occur at any time after the procedure.

The VARC-consensus definitions of major clinical endpoints reflect an orderly multi-step thought process and share multiple facets [3, 4, 5, 12].

1) Endpoints should include a specific set of criteria to evaluate device efficacy/safety and to gain insight into the biological effect of the new device. Understanding the biology of new devices is critical for moving innovative ideas forward in clinical practice.

  1. Clinicopathological correlation is the hallmark of modern medical diagnosis. Endpoints in cardiovascular intervention trials should relate to the pathophysiological mechanism(s) most likely responsible for the clinical outcome.
  2. Trial programmes that are appropriately designed will include the spectrum of patients presenting in routine practice and provide sufficient long-term follow-up of an adequate number of patients to adequately assess safety and efficacy. Clinical endpoints must assess procedure and device related risks and benefits (device/procedure related effectiveness and safety) as well as outcomes that are of importance to patients such as being able to live longer without symptoms, avoid rehospitalisations, and repeat procedures (patient-related effectiveness and safety). Figure 1 illustrates the case of a patient as opposed to a device oriented clinical endpoint at 2 years for the Resolute-All comers trial [13].

2) As emphasised earlier, the criteria for endpoint definition should allow for complete and timely collection and accurate assessment of the required clinical data. Outcome data provided should be sufficiently specific to be subjected to statistical analysis.

3) Clinical endpoints should provide a balance between the need to bridge back to historical data from previous percutaneous cardiovascular intervention/surgical trials and the need to implement definitions that reflect newly emerging knowledge in a global context, as well as unique and evolving aspects of the new technology. The endpoints must impart perspectives on already accepted endpoint definitions.

FOCUS BOX 2Operational definition of a consensus endpoint
Consensus endpoint definitions should enable insight into the biological effect of the new device while providing sufficient knowledge about what really matters – how patients fare with regard to freedom from all-cause death, myocardial infarction, and repeat revascularisation procedures

Consensus on definitions: The current definitions and beyond

Annex 3 lists the current ARC-2 [11], (M)VARC [4, 5], PARC [10], neuroARC [8], NARC [6], BARC [9], and HBR [7]-consensus endpoint definitions proposed [11] for uniform implementation in modern cardiovascular intervention trials. Details of the specific endpoint definitions were made public through peer-reviewed publication. In the following sections, we will focus on future challenges and on those elements where consensus is criticised, debated or in need of further clarification.

THE ACADEMIC RESEARCH CONSORTIUM (ARC) ( Table 1 , appendix 1)

FOCUS BOX 3The Academic Research Consortium
  • An informal collaboration between academia, industry and regulators to standardise the definition of trial endpoints
  • Currently consists of US (based at Harvard, Duke and the Cardiovascular Clinical Research Institutes), and European Investigators (based in Rotterdam, Bern and Paris), industry representatives and the US Food and Drug Administration
  • First convened in 2006
  • The ARC consensus endpoint definitions were initially conceived and work well for stable coronary disease patients with de novo lesions

The initial effort of the Academic Research Consortium (ARC) focused on establishing consistency among endpoint definitions and providing consensus recommendations for stent investigations in stable patients with de novo coronary lesions. That effort has now been expanded towards non-selected (study) patient populations. The latter, the all-comers study concept, may have important implications for consistency in implementation of the definitions and some aspects merit further discussion and clarification.

Death ( Table 1 , appendix 1)

ARC considers all-cause mortality the most unbiased method for reporting deaths in a clinical trial, even though it may be less specific than deaths adjudicated as due to cardiac aetiologies. For times when attribution to cardiac versus non-cardiac causes is desired, such as medium to long-term follow up studies, ARC proposed a conservative approach. Sudden and/or unexpected death, even in patients with coexisting potentially fatal non cardiac disease status (e.g., cancer, sepsis), should be classified as cardiac, including the unknown, unless history related to the non-cardiac diagnosis indicates death was imminent. Retrieval of autopsy reports offers a method of elucidating the true nature of death in clinical trials.

FOCUS BOX 4Reporting mortality endpoints
  • The ARC considers all-cause mortality the most unbiased method for reporting death
  • ARC-2 proposes that the cardiac and vascular categories be merged into a single entity, cardiovascular, in order to be consistent with pharmacological studies
  • Even when a distinction must be drawn between cardiac and non-cardiac mortality, the initial assumption is that a death is cardiovascular unless unequivocally demonstrated otherwise
  • Clinical autopsies may help to establish the true cause of death and should be strongly recommended

Myocardial infarction and the preferred biomarker of myocardial necrosis ( Table 2a , appendix 1)

MI may be an entry criterion for a clinical DES/BRS trial, may occur in the periprocedural period, or may occur long after device implantation as a result of spontaneous MI or late complications of the study device [14, 15].

The diagnosis of acute, evolving, or recent MI requires (in the absence of pathological confirmation) findings of a typical rise and/or fall of a biomarker of necrosis, in conjunction with clinical evidence (symptoms, or ECG) that the cause of myocardial injury is ischaemia [16].

MI definitions include certain arbitrary assumptions and vary across PCI, CABG, structural, and cardiovascular drug trials. In particular, clinical trial definitions of periprocedural (type 4a) MI have been difficult to standardize. The complexity stems from the use of increasingly sensitive biomarkers of subtle myocardial injury balanced against a need to identify events that clearly affect clinical outcomes, including mortality [11].

The use of CK-MB versus cTn has been especially controversial in the classification of periprocedural MI. The initial ARC definitions were designed for investigations in stable coronary disease patients with de novo lesions and proposed creatinine kinase myoband (CKMB) as the preferred marker of myocardial injury [1]. The expanded ability to detect myocardial injury using very sensitive and specific biomarker assays has impacted the Universal Re-definition of MI ( Table 2a ) and has critical consequences for the analysis and reporting of safety and efficacy data in contemporary cardiovascular (device) trials, most in particular when addressing myocardial infarction associated with cardiovascular interventions [16, 17, 18, 19]. This is of concern when addressing non-selected all-comers populations, including those with an on-going acute MI and other acute coronary syndromes [13, 20, 21, 22, 23].

Moreover, cTn assays pose several challenges from a clinical trial perspective. The analytic sensitivity among marketed cTn assays varies over several orders of magnitude. Therefore, the absolute values from 1 assay are not interchangeable with those from another. Furthermore, a several-fold increase (e.g., 5 URL) with a high-sensitivity cTn (hs-cTn) assay may not be associated with a detectable increase with a less sensitive earlier-generation assay, leading to significant differences in the event frequency based solely on the assay used [11].

The term high sensitivity reflects the characteristics of the assay, not a difference in the form of measured cTn. Specific criteria have been proposed by the International Federation of Clinical Chemistry and Laboratory Medicine to define hs-cTn assays [23]. Briefly, such assays should detect measurable levels of cTn in many (>50%) normal subjects and reliably detect changes (total imprecision, i.e., percent coefficient of variation <10%) at the 99th percentile URL.

ARC-2 endorses the classification of spontaneous MI and MI related to complications of the study device (types 1, 2, 3, 4b, and 4c) proposed by the 2018 universal definition of MI [11, 16].

For periprocedural MI and myocardial (re-)infarction/extension after PCI (Type 4 a), it will be important to distinguish events defined by a threshold level of enzyme or biomarker elevation where the degree of elevation has a proven relationship to other more meaningful clinical outcomes [24, 25]. Today, there are no data assigning clear thresholds for periprocedural troponin elevations and long-term prognosis following PCI, cardiac surgery, and other cardiovascular interventions. In variance with the 4th Universal definition of MI, ARC-II proposes a common ≥35 URL threshold for cTn for both PCI- and CABG -related periprocedural MI as a reasonable threshold [11, 16]. ARC proposes that 1 ancillary criterion ( Table 2b) be required in addition to the ≥35 cTn rise to fulfill the definition of periprocedural MI. The ancillary criteria are 1 or more of the following: “flow-limiting” angiographic complications in a major epicardial vessel or >1.5-mm-diameter branch, new significant Q waves (or equivalent) related to the procedure, or a “substantial” new wall motion abnormality on echocardiography related to the procedure.

ARC-II advocate that all data for (hs)cTn and CK-MB should be tabulated for each classification to include at least the following multiples of the URL by treatment groups: >1, ≥5, ≥10, ≥35, ≥70, ≥100, and ≥250. Cumulative frequency distributions for cTn and CK-MB by treatment group should be provided [11]. This absolute rise applies both to patients with baseline cTn levels <URL and to those in whom baseline cTn levels are elevated and stable or falling [11].

In Table 2a, appendix 1 shows the updated ARC2 consensus definitions for all commers patients, and for comparison the 2018 universal definition of MI criteria and the Society for Cardiac Angiography and Interventions criteria.

Device related complications: revascularisation ( Table 3a , appendix 1)

ARC-II defined target lesion revascularisation (TLR) ( Table 3a): Any repeat PCI to the target lesion including 5 mm proximal and distal from the edge of the stent, or bypass surgery of the target vessel for clinically significant re-narrowing or other complication of the target lesion. In variance to the initial ARC definition, priority was given to functional assessment with fractional flow reserve (FFR) ( Table 3b) or equivalent techniques over an independent quantitative coronary angiography (QCA) assessment at the target lesion or in the target vessel ( independent core laboratory) [11].

In a real world setting, considering patient comorbidity (e.g., renal function) and/or the complexity of the coronary lesions leading to myocardial ischaemia, the investigators may also decide to stage the coronary revascularisation involving two or more procedures. Staged procedures are considered part of the index procedure.

Criteria for staging have to be well defined [26]. Staging cannot involve the target lesion or any adjacent branches. Some protocols will exclude any revascularisation involving the target vessel from the staging process.

Device related complications: stent or scaffold thrombosis ( Table 4 , appendix 1)

The biohazard of late stent or scaffold thrombosis is a catastrophic complication of coronary stenting, presenting as sudden death, non-fatal myocardial infarction and/or the need for target lesion revascularisation in almost all cases. The impact of individual components of the proposed patient-oriented clinical endpoint of death, non-fatal myocardial infarction and repeat revascularisation procedures on other safety outcomes, such as stent thrombosis, remains difficult to measure. Major concern remained, surrounding events occurring beyond one year that may present as unexplained death.

The ARC classification for ST categorised the clinical endpoint into definite, probable, and possible, as well as acute (0-24 hours), subacute (>24hours-30 days), late (>30days-1year), and very late (>1year). Early stent thrombosis is acute and subacute stent thrombosis (0-30days). The initial ARC stent thrombosis classification included possible stent thrombosis to capture potential stent thrombosis events occurring beyond 1 year that may present as unexplained death [1]. However, it has become apparent that an increasing proportion of deaths remain unexplained over time, leading to a potential exaggeration of the rates of very late stent thrombosis and possible dilution of a signal for differences between groups resulting from noise created by inclusion of events unrelated to stent thrombosis [27].

The ARC definite or probable categories have emerged as the preferred method for assessing stent thrombosis based on concerns for exaggeration of stent thrombosis rates, but it is clear this may instead underestimate the true rate as it does not account for late or very late events presenting as death [22, 28]. This classification included unexplained death within 30 days after the procedure as a criterion for probable stent thrombosis. Given the uncertain specificity of this criterion after ST-segment–elevation MI, the updated ARC-2 classification recommends no longer including unexplained death within 30 days as a criterion for probable stent thrombosis when the index procedure was performed in the context of ST-segment–elevation MI, albeit with evident implications for historical comparisons [11].

Moreover, ARC-2 recommends intraprocedural stent thrombosis as an additional, optional classification but does not include it as a subcategory of stent thrombosis, the timing of which begins when the patient has been taken off the catheterization laboratory table [11].

Patient reported outcomes

Patient-reported outcomes systematically measure health status directly from the patient, without amendment or interpretation of the response by a clinician or anyone else. Patient-reported outcomes should be routinely considered in the design of clinical trials to provide important information for subsequent medical decision making and shared medical decision making with patients.

The Neurologic Academic Research Consortium (NeuroARC)

Surgical and catheter-based cardiovascular procedures and adjunctive pharmacology have an inherent risk of neurological complications. Our ability to interpret the risks associated with procedure-related neurovascular injury is challenged by the lack of a conclusive link between acute procedure-related subclinical brain lesions and long-term neurological or cognitive outcomes.

NeuroARC aimed to define the full spectrum of neurovascular injury post procedure, with the assumption that standardized data acquisition will accelerate differentiation between clinically meaningful and incidental findings[8]. However, NeuroARC does not address acute stroke interventions, which have distinct therapeutic considerations.

Brain injury related to cardiovascular procedures spans a spectrum from overt stroke to covert injury, and can be classified according to clinical signs and symptoms and neuroimaging ( Table 5a , appendix 2) [8].

Stroke is among the most feared complications of surgical and transcatheter cardiovascular interventions, affecting both benefit-risk evaluations and health care costs. Clinical manifestations of periprocedural stroke are highly variable and substantially under-reported [8]. Stroke is the acute onset of symptoms consistent with focal or multifocal CNS injury caused by vascular blockage resulting in ischaemia or vascular rupture resulting in haemorrhage (infarction with haemorrhagic conversion, with or without mass effect), and is distinct from global hypoxic-ischaemic injury, and should be reported separately. Also, multifactorial, delirium (global neurological dysfunction) without CNS injury should be reported due to its prognostic implications [29, 30]. Central Neural System (CNS) haemorrhage should be classified as a stroke when it is not caused by trauma, and needs confirmation by imaging; major types include intracerebral haemorrhage and subarachnoid haemorrhage.

The proposed standardized neurological endpoints for cardiovascular clinical trials, and the recommended timing of clinical (i.e., acute severity, recovery, and long term disability; Table 5b) and imaging evaluations, are depicted in table and figure . NeuroARC strongly recommends cognitive screening for all cardiovascular trials, and a comprehensive cognitive assessment strategy for studies with neurological outcomes as efficacy endpoints [8]. Magnetic Resonance Imaging (MRI) is the imaging modality of choice for detection and quantification of brain ischemia related to cardiovascular procedures [8].

Peripheral Academic Research Consortium (PARC)

Lower extremity peripheral artery disease (LE-PAD) is a manifestation of systemic atherosclerotic disease [31, 32, 33]. Although LE-PAD can be asymptomatic and subclinical, it is associated with a reduction in functional capacity and quality of life when symptomatic, and, in its most severe form, is a major cause of limb amputation [31]. Patients with LE-PAD are at an increased risk for myocardial infarction (MI), stroke, and death [31,34-37).

The PARC group faced fundamental challenges in creating broadly accepted, pragmatic LE-PAD definitions, including the scope of topics requiring definitions [10]. Across this clinical spectrum, definitions for patients were required that included the accurate elucidation of symptoms, anatomic characterization of disease, definitions for both clinical and imaging short- and long-term measures, and finally, clinical outcomes. In addition to developing definitions applicable across the wide spectrum of PAD syndromes, the definitions would need to be pertinent to existing and developing therapies and procedures for LE-PAD [10].

The PARC group established descriptive (rather than numeric) baseline symptom definitions benchmarked to the established Fontaine and the Rutherford classification systems [10, 38]( Table 6 , appendix 3)

The PARC lesion and vessel definitions are presented in Table 7 (appendix 3). In defining a “significant” anatomic lesion in the LE-PAD arterial tree, PARC considered options similar to those evaluated in the coronary circulation, including classification within the 50% to 100% stenosis/occluded group [10].

Acute technical success for peripheral revascularization is defined as the achievement of a final residual diameter stenosis <30% for stent and <50% for angioplasty or atherectomy by angiography at the end of the procedure (and without flow-limiting arterial dissection or hemodynamically significant trans-lesional pressure gradient <10 mmHg) for endovascular revascularization or patent vessel or bypass conduit for surgical revascularization ( Table 8 , appendix 3). Acute procedural success for peripheral revascularization is defined as both acute technical success and absence of major adverse events (e.g., death, stroke, MI, acute onset of limb ischemia, index bypass graft or treated segment thrombosis, and/or need for urgent/emergent vascular surgery) within 72 h of the index procedure. Clinically driven target vessel revascularization and vessel patency were defined. Short- and long-term surrogate endpoints for procedural success used imaging and physiologic measures at 72 hours to 30 days post index procedure to 3, 6, and 12 months ( Table 8 , appendix 3).

PARC Functional/clinical outcome definitions for patients with intermittent claudication and critical limb ischemia are shown in Table 9 (appendix 3). Major adverse limb events were defined as an above-ankle amputation of the index limb or major repeat revascularization of the target limb (new bypass graft, jump/interposition graft revision, repeat endovascular therapy, or thrombectomy/thrombolysis). The PARC major adverse limb event (MALE) definition includes all repeat open and endovascular interventions in the target limb. Early wound healing was defined as complete epithelialization of an ischemic wound of the target limb that stayed closed for at least 2 weeks, evaluated at 30 days.

THE VALVE ACADEMIC RESEARCH CONSORTIUM (VARC) ( Table 10 , appendix 4 )

View chapter on TAVI

Transcatheter Heart Valve related therapy is a fast emerging technology. Since the introduction of Transcatheter Aortic Valve Implantation (TAVI) to the medical community in 2002, there has been increasing interest in the field of this catheter based, lesser invasive, treatment in many patients with aortic stenosis. The development and implementation of this radically new treatment should follow the standard bed-to-bedside evidence based medicine pathway, starting with pre-clinical bench testing, and advancing to meaningful clinical endpoint trials. In the spirit of the ARC-mission statement [2], the VARC convened and responded to the urgent need for standardised clinical research processes for transcatheter aortic valve implantation trials.

As indicated, the priority and responsibility of the VARC consortium was directed towards, and limited to, specifying and standardising definitions for clinical endpoints of TAVI trials, focusing on patients at high surgical risk. The inter-disciplinary nature of TAVI, combining aspects of both surgical and interventional therapies, presented special challenges and required an enlightened and collaborative approach to the development of clinical research recommendations and endpoint definitions. The agreed upon clinical endpoints strike a delicate balance encompassing features of interventional clinical trial methodologies and historical surgical perspectives on prosthetic valve clinical and hemodynamic performance ( Appendix 4 , Table 10 ) Consensus about patient selection, the use of surgical risk algorithms to identify high or prohibitive risk surgical patients; trial design and statistical analyses for valve-related clinical trials (e.g., superiority vs. non-inferiority analysis) were considered important, however not the primary subject of the parent manuscript. They should be addressed in dedicated follow up consensus documents. Indeed, the VARC consensus definitions should be considered as a first step and may also provide an opportunity for extension towards other Transcatheter Heart Valve Therapy. In most cases, endpoint definitions for TAVI are characterised in relation to the specific implant device, the implant procedure and the resultant patient-oriented outcomes assessed longitudinally for the life of the implant. VARC adopted a more centrist approach to MI definitions after TAVI; and many global endpoints such as stroke, bleeding, and some specific endpoints (e.g., paravalvular prosthetic regurgitation) are themselves in a state of evolution, subject to other consensus committees in the near future. The maturation of technology and broadening experience with this therapy mean that certain definitions have become ambiguous and need of updating and clarification. Indeed, the definitions proposed in VARC were re-appraised and updated in 2012 and termed VARC-2

with definitions covering the following areas [4]:

  1. Risk factors not captured in traditional scoring systems
  2. The role of the Heart team ( View dedicated chapter )
  3. Mortality
  4. Myocardial infaction
  5. Stroke
  6. Bleeding complications (no change from VARC)
  7. Acute Kidney Injury
  8. Vascular complications
  9. Conduction disturbances and arrhythmias
  10. Other TAVI-related complications
  11. Valvular function
  12. Quality of life
  13. Composite endpoints

Traditional risk scores have limited applicability for the group of patients who are deemed inoperable and yet may be candidates for TAVI. Characteristics such as porcelain or severely atherosclrotic aorta, frailty, severe liver disease, anatomical variables that make stenotomy or re-sternotomy hazardous, and the health of the right heart all enhance the ability to risk stratify this group of patients.

A second new area explored and recommended by VARC-2 was the use of the heart team for patient evaluation. This mirrors guidelines from the ESC for the management of valvular heart disease [4] ( View 'The Heart Team' chapter ).

The heart team should include as a minimum interventional cardiologist, cardiothoracic surgeons, and imaging specialists. It, however, should be dynamic and utilise the expertise of other specialists such as anaesthesiologists and geriatricians when appropriate. Regular scheduled meetings are recommended to allow for the adjustment of decision-making according to local experiences and resources as well as to enable customised decisions for unusual or complex clinical situations (such as concomitant coronary disease), patient selection, procedural performance, and the management of complications.

The heart team needs to arrive at an agreement on an estimated 30-day mortality for each patient under consideration. Despite the challenge of assigning limits to surgical risk strata the following are recommended:

Low risk: <4%

Intermediate risk: 4-10%

High risk: >11-15%

Very high risk: 15-50%

Extreme risk (inoperable): >50% (mortality or irreversible morbidity) as judged by at least two cardiothoracic surgeons from a tertiary centre specialising in SAVR.

In addition to the selection of inoperable or high risk patients for TAVI the heart team should agree on those in whom any invasive therapy would be futile (TAVI or SAVR). Life-expectancy indices and/or expected quality of life improvement play a key role in these decisions.

Mortality

VARC proposed procedural all-cause mortality, cardiovascular mortality, and non-cardiovascular mortality data to be collected within 30 days or during the index hospitalization (if longer than 30 days). In addition, VARC-2 proposes a sub-category of immediate procedural mortality within 72 hours.

Myocardial infarction and the preferred biomarker of myocardial necrosis

VARC proposed to define periprocedural MI as an acute event that is associated with documented and >20% increase in the second, post procedure sample of CKMB and above ten times the upper normal range. The periprocedural interval is inclusive of all events that begin within 72 hours of the index procedure. Of importance, since cTn measurements prove an extremely sensitive biomarker of myocardial necrosis, and since we lack actual data assigning clear thresholds for periprocedural cTn elevations and long-term prognosis after either PCI or cardiac surgery (See also ARC II recommendations), VARC did not consider cTn as the periprocedural biomarker of choice in TAVI clinical trials at the time. Nevertheless, troponin measurements should be obtained and reported for future comparative analyses.

Following VARC-2, however, the thresholds and definitions have been adjusted. Troponin >10 times the upper limit or CK-MB >5 times the upper limit <72hours from the procedure, in combination with the same clinical criteria, are considered sufficient to identify periprocedural myocardial infarction. To enhance detection of infarction the systematic collection of biomarkers at within 12-24hours, at 24 hours, at 72 hours or at discharge, and, if elevated daily until a decline is seen, is recommended.

FOCUS BOX 5Reporting acute myocardial infarction
  • MI cannot be defined solely by biomarker elevations in the absence of new clinical signs (e.g., new pathological Q-wave) or symptoms consistent with myocardial ischaemia
  • Acute ischaemic events occurring after 72 hours, and after the cardiac biomarkers have returned to normal, are considered spontaneous MIs and are defined in accordance with the universal MI guidelines

Stroke

Strokes during and after TAVI may occur due to embolic events from multiple sources (aorta, left ventricle, native or prosthetic valve), procedure-related aortic dissections, ischaemia from hypotension during the procedure, or haemorrhagic complications associated with adjunctive pharmacotherapy. VARC considered five important elements in arriving at clinically relevant stroke event definitions without creating undue bias in favour of transcatheter therapies. They include:

  1. a clinical diagnosis of stroke which ruled out metabolic or toxic encephalopathy, pharmacological influences, and non-central neurologic symptoms,
  2. the role of neuro-imaging studies (preferably diffusion-weighted MRI brain imaging) for confirmation of the diagnosis,
  3. the distinction of stroke versus transient ischaemic attack (TIA) (including timing),
  4. categorisation of stroke into major and minor events based on the degree of disability as defined by conventional neurological assessment tools, and
  5. sub-classification of strokes into haemorrhagic, ischaemic, and undetermined categories. The latter stratification aligns with the newly proposed definitions by a recent FDA consensus panel. Ischaemic stroke is as an acute symptomatic episode of focal cerebral, spinal, or retinal dysfunction caused by an infarction of central nervous system tissue. Haemorrhagic stroke is an acute symptomatic episode of focal or global cerebral or spinal dysfunction caused by a non-traumatic intraparenchymal, intraventricular, or subarachnoid haemorrhage. An undetermined stroke is a stroke with insufficient information to allow categorisation as either of ischaemic or haemorrhagic origin. Stroke severity is classified upon symptom severity and persistent patient disability graded by the modified Rankin score and National Institutes of health Stroke Scale (NIHSS).

To further align with the FDA consensus the VARC-2 consensus has refined the earlier VARC definitions. Major or minor stroke are replaced by Diabling and non-disabling. The former is one that results (at 90 days after stroke onset) in an mRS score of ≥2 and an increase in ≥1 mRS category from an individual’s pre-stroke baseline. The latter is one that results (at 90 days after stroke onset) in an mRS score of 2 or that does not result in an increase in ≥1 mRS category from an individual’s pre-stroke baseline. In addition to this categorization of disabling and non-disabling strokes, the endpoint of all strokes should be reported.

Bleeding

The VARC definition of clinically meaningful bleeding was guided by the following principles: [1] the definition must be based on objective criteria, including an obvious source of bleeding or number of transfusions while excluding bleeding from “obscure origin”; [2] serious or meaningful bleeding must result in death, be life-threatening, be proven to be associated with increased long-term mortality, cause chronic sequelae, or consume major health-care resources. Yet, given the ample body of literature suggesting that administration of whole blood or RBC transfusions in patients with cardiovascular pathology may be potentially harmful, any whole blood or RBC transfusion needs to be reported in the case report forms, including the number of transfused units, regardless of the presence or absence of overt bleeding. The VARC-2 consensus maintains the BARC definitions, upon which VARC bleeding definitions were based, but stresses that bleeding complication has to be the result of overt bleeding and cannot be adjudicated by the number of RBC transfusions alone.

Renal dysfunction

Acute Kidney Injury (AKI) independently has been associated with post procedure mortality in patients undergoing TAVI [39].In defining the stages of AKI, VARC adopted the modified RIFLE (Risk, Injury, Failure, Loss and End-stage Kidney Disease) classification, however, focussing only on potential changes in creatinine level. AKI was defined by either RIFLE stage 2 or 3 [40]. An outer bound of 72 hours from the index procedure for diagnosing AKI was selected based on evidence that adverse outcomes with small changes in serum creatinine were observed when the elevation occurred within 24 to 48 hours of the procedure and to ensure that the process was both acute and related to the procedure itself rather than as a consequence of post-procedure end-stage multi-organ system failure. However, any case of renal replacement therapy in hospital and up to 30 days should be reported. The VARC-2 criteria have replaced the RIFLE criteria with the AKIN criteria [40].

FOCUS BOX 6Reporting acute kidney injury for TAVI trials
  • VARC favoured the RIFLE criteria for classifying AKI. VARC-2 favour the AKIN criteria
  • An outer bound of 72 hours from the index procedure for diagnosing AKI was proposed by VARC. This has been extended to 7 days in VARC-2
  • Any use of renal replacement therapies within 30 days of the index procedure should be reported

Procedure-related (specific) complications

Prosthetic aortic valve procedures, and in particular TAVI, may have untoward effects on the aortic valvular complex, that may result in important clinical consequences. Therefore, VARC proposed to group these complications as a separate endpoint category. It must be noted, however, that some of the adverse events listed in this section may or may not be directly related to the valve prosthesis. They may occur, for example, during balloon aortic valvuloplasty (e.g., conduction disturbances and cardiac arrhythmias). This approach in VARC has been updated in VARC-2 to provide precise definitions of complications not captured by other categories.

Access site complications

The access site was defined as any location (arterial or venous) traversed by a guidewire, a catheter, or a sheath (including the LV apex and the aorta) and access related is defined as any adverse clinical consequence possibly associated with any of the access sites used during the procedure. The time course for vascular complications is usually during the index hospitalisation, but occasionally a procedure-related vascular complication may have important clinical manifestations (early or late) as an outpatient (e.g., ruptured pseudo-aneurysm [femoral or left ventricle] or progressive limb ischaemia).

In defining vascular complications, VARC made reference to the reporting standards of the Society of Vascular Surgery for defining and reporting vascular complications following endovascular aortic graft repair procedures [41].

Of importance, femoral vascular access and closure in many centres is routinely achieved using surgical cut-down procedures, and therefore, a documented pre-procedural commitment to a surgical access/closure should be considered as part of the procedure and not as a complication.

VARC-2 has clarified the definitions so that pre-planned surgical or endovascular access should be considered part of the procedure and not a complication unless there are untoward clinical consequences. Unplanned endovascular stenting or surgical repair without clinical sequelae are to be considered a minor vascular complication.

Prosthetic valve performance

VARC proposed two criteria to evaluate prosthetic valve performance; 1) prosthetic valve haemodynamics assessed on repetitive basis by echocardiography and 2) associated clinical findings indicating impaired cardiovascular or valvular function. In VARC-2 these definitions were refined to take account of larger than average or smaller than average body habitus. ( Appendix 4 , Table 10 )

Quality of life

VARC-2 proposes that health-related quality of life assessments (heart failure specific and generic) be used at multiple time-points. Early assessment at 2 weeks, 1 month and 3 months with intermediate

Composite endpoints

The VARC consensus recommended device success, early safety, and clinical efficacy as composite endpoints. VARC-2 extends this list to include the category of time-related valve safety, which combines valve dysfunction, endocarditis, and thrombotic complications of the prosthesis.

THE mitral valve ACADEMIC RESEARCH CONSORTIUM (MVARC) ( Table 11 , appendix 5 )

Mitral regurgitation (MR) is one of the most prevalent valve disorders and has numerous etiologies, including primary (organic) MR, due to underlying degenerative/structural mitral valve (MV) pathology, and secondary (functional) MR, which is principally caused by global or regional left ventricular remodeling and/or severe left atrial dilation.

Diagnosis and optimal management of MR requires integration of valve disease and heart failure specialists, MV cardiac surgeons, interventional cardiologists with expertise in structural heart disease, and imaging experts.

The introduction of transcatheter MV therapies has highlighted the need for a consensus approach to pragmatic clinical trial design and uniform endpoint definitions to evaluate outcomes in patients with MR.

Where possible, MVARC has endeavored to align their consensus definitions with other ARC initiatives, and other professional society and organization efforts (with greater granularity, when necessary, specific to MR therapies), while incorporating the latest knowledge from clinical studies. Table 11 contains the list of endpoints relevant to mitral interventions. Other important secondary endpoints, including quality-of-life measures, functional performance, and echocardiographic assessments [12].

MVARC recommends that the basic definitions (classifications, and levels of severity) for death, bleeding, renal dysfunction, and neurologic events, after MV procedures generally conform to VARC-2 criteria. For periprocedural myocardial infarction after MV procedures the SCAI criteria for clinically relevant periprocedural MI are proposed [42]. However, MVARC acknowledges that the implications of biomarker elevations after surgical incisions of the myocardium (including transapical access) are unknown, and that at present, a strict definition of MI in valve procedures should require additional evidence of myocardial injury, either new ST-segment elevation or depression or pathological Q waves[5]. For non-periprocedural spontaneous MI, the Third Universal Definition of Myocardial Infarction criteria are endorsed [17].

Vascular access site complications

Transcatheter MV interventions may result in complications arising from access site entry, trans-septal procedures, and/or the device interacting with the MV complex and adjacent structures ( Table 11). Vascular complications directly due to malfunction of closure devices should also be recorded. Planned repair of access site entry portals such as the myocardial apex are not considered access site-related complications. Although atrial septal defects after trans-atrial procedures are not usually of hemodynamic significance and do not require repair, the criteria for a significant residual atrial septal defect should be pre-specified.

Arrhythmias and Conduction System Disturbances

Assessment of the presence and burden of atrial and ventricular arrhythmias, conduction system disease, and use of therapeutic drugs and technologies is important in this patient population. Pacemaker or defibrillator lead dislodgement is also a potential risk, especially if the leads were recently positioned, and should be reported.

Specific Device-Related Technical Failure Issues and Complications

Device-related specific endpoints include subcategories of device access and delivery-related success and complications, acute and chronic device function, and device-related complications. The endpoints presented in 5 represent a framework for device-related endpoints that should be considered, adjusted as appropriate for each particular device.

Composite endpoints

Technical, device, procedural and patient success are complementary metrics used to evaluate the acute and late outcomes of MV surgical and transcatheter interventions.

THE BLEEDING ACADEMIC RESEARCH CONSORTIUM (BARC) ( Table 12 , appendix 6 )

View chapter on 'Peri-procedural and post-procedural antithrombotic pharmacotherapy'

The standardised bleeding consensus definitions for cardiovascular trials are a welcome extension to the ARC consensus definitions. Bleeding complications have been established as an important safety endpoint for cardiovascular, pharmacologic as well as cardiovascular device, trials. They have been associated with an increased risk of subsequent adverse outcomes, including MI, stroke, and death [43, 44]. The exact mechanisms underlying this relationship are not known but may include cessation of evidence-based cardioprotective agents, including antiplatelet agents, the direct effects of blood transfusion, as well as the deleterious role of anaemia.

There is a diversity of bleeding event definitions available in literature at this time. They incorporate a different combination of both laboratory parameters and clinical events and rank order these combinations into difference severity categories. In their effort to create a universal consensus bleeding classification, the BARC faced multiple challenges ( Table 6 ). The proposed BARC bleeding classification was conceived to capture information about the aetiology (procedural or nonprocedural), the site of bleeding (intraocular, intracranial, visceral, peritoneal, access site, etc.), and the severity of bleeding (quantified by impact on laboratory data as well as clinical status). CABG related bleeding was integrated in the proposal [9]. It was intended to strike an appropriate balance between sensitivity and specificity of criteria, in order to optimise the ability to detect dose response or discern small variations between therapies, while likewise arriving at clinically meaningful conclusions. The proposed BARC bleeding classification may be considered a composite, consisting of multiple, hierarchically graded categories of severity with different relation towards clinical outcome.

Academic Research Consortium for High Bleeding Risk (HBR ARC)

Identification and management of patients at high bleeding risk undergoing percutaneous coronary intervention are of major importance. The 2017 European Society of Cardiology focused update on DAPT in coronary artery disease recommended (Class IIb recommendation, Level of Evidence A) that the use of risk scores such as the PRECISE-DAPT (Predicting Bleeding Complications In Patients Undergoing Stent Implantation and Subsequent Dual Anti Platelet Therapy) and DAPT scores may be considered to guide antiplatelet therapy after PCI among others [45] [46, 47]. However, the scores proposed, among others had only moderate accuracy for predicting bleeding, with C statistics in the development cohorts ranging from 0.64 to 0.73[46, 47, 48, 49, 50, 51, 52]. None of these scores was validated in HBR patient populations.

By consensus of the HBR ARC group, HBR was defined as a BARC 3 or 5 bleeding risk of ≥4% at 1 year or a risk of an intracranial haemorrhage (ICH) of ≥1% at 1 year[7]. Thus, a major criterion for ARC-HBR is defined as any criterion that, in isolation, is considered to confer a BARC 3 or 5 bleeding risk of ≥4% at 1 year or any criterion considered to be associated with a risk of ICH of ≥1% at 1 year. A minor criterion for ARC-HBR is defined as any criterion that, in isolation, is considered to confer increased bleeding risk, with a BARC 3 or 5 bleeding rate of <4% at 1 year. Patients are considered to be at HBR if at least 1 major or 2 minor criteria are met ( Table 13, appendix 7). All major and the majority of minor ARC-HBR criteria identify in isolation patients at HBR [48, 49].

Non-adherence Academic Research Consortium (NARC)

Non-adherence to pharmacological therapy even in the controlled environment of clinical trials has been well recognized for years and has complex and potentially major effects both on trials’ results and interpretation [50, 51].

Optimal adherence in the context of a clinical trial comprises initiation (a binary variable), implementation (adherence to medication based on dosing history), and persistence until recommended discontinuation. While medication initiation and discontinuation pose relatively minor difficulties, the challenges derived from the complex nature of implementation have been well documented as have the shortcomings of traditional methods (mostly based on patient self-reporting or returned pill counts) of measurement [52].

Based on the aforementioned principles, a four-level classification of non-adherence was proposed, capturing a gradient of non-adherence, in a hierarchical fashion, from initiation through implementation to the discontinuation of treatment [6]. The severity of non-adherence patterns is standardized based on the expected degree of over- or more frequently under-exposure to study medication for any given experimental drug.

Three additional layers of information characterize the decision-process and circumstances underlying non-adherence, the clinical scenario and, where relevant, the timing relative to treatment initiation [6]. The pharmacological life is used as the discriminator to clearly distinguish different scenarios considering residual pharmacological activity of the study drug.

The NARC accepts a 5% tolerance as a reasonable threshold to define optimal adherence [6]. The RESULT acronym has been developed to help memorize and easily categorize the most frequent clinical scenarios resulting in non-adherence ( Table 14).

Conclusions

Although consensus criteria will inevitably include certain arbitrary features, consensus criteria for clinical end points provide consistency across studies, registries, or even data base analysis that can facilitate the evaluation of safety and effectiveness of drugs and devices.

Personal perspective – Pascal Vranckx

The critical challenge in the conduct of endpoint trials relates to the definition, collection and accurate assessment of endpoint data in a consistent timely manner. The success of a clinical endpoint study depends on understanding the challenges and incorporating the special requirements of these studies into the protocol design and operational procedures throughout the study.
In principle, the consensus process calls for the highest standards of clinical research, including (1) determination and implementation of consensus definitions (subject of this manuscript) (2) harmonisation and structured organisation of data collection (3) interpretation and reporting of events, and (4) the consistent use of central core laboratories and independent, blinded adjudication.
Consensus definitions are dynamic by nature; yet future modifications should follow a similar mechanism for consensus, respect the established historical definitions, whilst aiming to address potential limitations. Updates are required for the current ARC definitions, aiming at real world populations including complex lesions and high-risk populations (both ST segment and non-ST-segment elevation ACS), especially considering the current evolution in the field of cardiac biomarkers.

The VARC definitions focused on patients at high surgical risk, but implementation in lower risk patient populations should be envisaged. The VARC consensus definitions are conceived to be sufficiently broad to allow expansion to other areas in catheter based valve technology (mitral valve, tricuspid valve disease).

SHARE YOUR COMMENT