Updated standardized endpoint definitions for transcatheter aortic valve implantation: the Valve Academic Research Consortium-2 consensus document†

This is not an experiment but a consensus-building process. The researchers tested nothing in the traditional sense. Instead, they convened experts to agree on standardised definitions for clinical endpoints in TAVI studies. The intervention was the consensus process itself: two in-person meetings, working groups, and iterative drafting. The "outcome" was a set of agreed-upon definitions for:

Mortality (all-cause, cardiovascular, and non-cardiovascular)

Stroke (disabling vs non-disabling, using modified Rankin Scale)

Myocardial infarction (peri-procedural vs spontaneous, with specific biomarker thresholds)

Bleeding complications (life-threatening, major, minor)

Acute kidney injury (stages 1–3 based on creatinine and urine output)

Vascular complications (major vs minor)

Conduction disturbances and arrhythmias (new pacemaker implantation, new-onset atrial fibrillation)

Prosthetic valve dysfunction (structural deterioration, non-structural dysfunction, thrombosis, endocarditis)

Quality of life assessments (using validated instruments like the Kansas City Cardiomyopathy Questionnaire and the SF-36)

The document also defines composite endpoints (e.g., "early safety at 30 days" combining all-cause mortality, stroke, life-threatening bleeding, acute kidney injury stage 3, coronary obstruction, major vascular complication, and valve-related dysfunction requiring re-intervention).

No patients were studied. The "subjects" were:

40–50 experts including interventional cardiologists, cardiac surgeons, imaging specialists, neurologists, geriatric specialists, clinical trialists

Representatives from the US Food and Drug Administration (FDA)

Industry representatives from device manufacturers

VARC study group members

The meetings occurred in September 2011 (Washington, DC) and February 2012 (Rotterdam, Netherlands). The document was published in 2012.

No instruments were used on patients. The "measurement" was the degree of consensus among experts. The process involved:

Literature review of existing endpoint definitions from prior TAVI trials and surgical aortic valve replacement studies

Review of the original VARC-1 definitions (published 2011) to identify ambiguities and gaps

Working groups assigned to each endpoint category, who drafted definitions

Full group discussion and voting to reach consensus

Iterative revision based on feedback from FDA and industry representatives

The document itself serves as the measurement tool—it provides explicit criteria for each endpoint. For example, stroke is defined using the modified Rankin Scale (mRS) score of 0–6, with disabling stroke defined as mRS ≥2 at 90 days. Bleeding is classified using the Bleeding Academic Research Consortium (BARC) criteria adapted for TAVI.

**Study design:** This is a consensus document, not a clinical trial. It follows a modified Delphi process—an iterative method for achieving expert agreement on complex topics. Two face-to-face meetings were held, with working groups conducting preparatory work between meetings.

**Randomisation:** None. This is not applicable to a consensus process.

**Blinding:** None. All participants knew each other's identities and affiliations. Industry representatives were present, which introduces potential bias.

**Duration:** The process spanned approximately 6 months (September 2011 to February 2012), with additional time for manuscript drafting and review.

**Statistical approach:** No statistical analysis was performed. Consensus was defined qualitatively as "substantive agreement" among participants, though no formal threshold (e.g., 75% agreement) was specified.

**What this design can and cannot prove:**

*What it can prove:*

Expert opinion on best practices for endpoint definition

Areas of agreement and disagreement among specialists

A framework for standardising outcome measurement in TAVI trials

*What it cannot prove:*

Whether these definitions are clinically valid or predict patient outcomes

Whether using these definitions changes trial results or patient care

Whether the definitions are superior to alternative approaches

Any causal relationships between TAVI and clinical outcomes

**Major methodological weaknesses:**

No systematic literature search or meta-analysis to ground definitions in evidence

Industry presence may have influenced definitions to favour device manufacturers

No patient representatives included in the consensus process

No formal voting thresholds or pre-specified criteria for consensus

The document reflects expert opinion from 2011–2012, which may be outdated given subsequent evidence

No validation of definitions against real-world data or prospective testing

The document produced standardised definitions for the following endpoints:

**Mortality:**

All-cause mortality: death from any cause

Cardiovascular mortality: death due to cardiac causes, vascular causes, or unknown cause

Non-cardiovascular mortality: death due to a clearly documented non-cardiovascular cause

Timing categories: procedural (≤72 hours), in-hospital, 30-day, 1-year, and annual thereafter

**Stroke:**

Modified Rankin Scale (mRS) score assessed at baseline and at 90 days

Disabling stroke: mRS ≥2 at 90 days (or increase of ≥1 from baseline if baseline mRS ≥2)

Non-disabling stroke: mRS <2 at 90 days

Transient ischaemic attack (TIA): neurological deficit lasting <24 hours with no imaging evidence of infarction

**Myocardial infarction (MI):**

Peri-procedural MI (≤72 hours after TAVI): cardiac troponin elevation >15× upper reference limit (URL) plus either new ischaemic ECG changes, new pathological Q waves, or angiographic evidence of new coronary occlusion

Spontaneous MI (>72 hours after TAVI): troponin elevation >URL plus symptoms, ECG changes, or imaging evidence

Periprocedural MI after surgical aortic valve replacement: CK-MB >10× URL plus new Q waves

**Bleeding:**

Life-threatening or disabling bleeding: fatal bleeding, bleeding into a critical organ (intracranial, intraocular, intraspinal, retroperitoneal, pericardial with tamponade), bleeding causing hypovolaemic shock, or bleeding requiring transfusion of ≥4 units of packed red blood cells

Major bleeding: overt bleeding with haemoglobin drop ≥3.0 g/dL, transfusion of 2–3 units, or bleeding requiring surgery

Minor bleeding: any overt bleeding not meeting major or life-threatening criteria

**Acute kidney injury (AKI):**

Stage 1: serum creatinine increase 1.5–1.9× baseline or ≥0.3 mg/dL increase, or urine output <0.5 mL/kg/hr for 6–12 hours

Stage 2: serum creatinine increase 2.0–2.9× baseline, or urine output <0.5 mL/kg/hr for ≥12 hours

Stage 3: serum creatinine increase ≥3.0× baseline, or ≥4.0 mg/dL increase, or initiation of renal replacement therapy, or urine output <0.3 mL/kg/hr for ≥24 hours, or anuria for ≥12 hours

**Vascular complications:**

Major: aortic dissection, aortic rupture, access site or access-related vascular injury leading to death, need for blood transfusion ≥4 units, or unplanned percutaneous or surgical intervention

Minor: access site or access-related vascular injury not meeting major criteria

**Conduction disturbances:**

New permanent pacemaker implantation within 30 days

New-onset left bundle branch block

New-onset atrial fibrillation

**Prosthetic valve dysfunction:**

Structural valve deterioration: permanent intrinsic changes of the valve (calcification, leaflet tear, flail) leading to stenosis or regurgitation

Non-structural dysfunction: paravalvular regurgitation, patient-prosthesis mismatch, or malposition

Valve thrombosis: confirmed by imaging or pathology

Endocarditis: definite or possible per modified Duke criteria

**Composite endpoints:**

Early safety at 30 days: all-cause mortality, all stroke, life-threatening bleeding, AKI stage 3, coronary obstruction requiring intervention, major vascular complication, valve-related dysfunction requiring re-intervention

Clinical efficacy at 1 year: all-cause mortality, all stroke, hospitalisation for valve-related symptoms or worsening heart failure, valve-related dysfunction requiring re-intervention

Since this is a consensus document with no experimental data, there are no effect sizes, confidence intervals, or p-values. The "effect" is the standardisation itself—the document provides a common language that allows comparison across studies. Before VARC-2, a "major bleeding" event in one trial might be classified as "minor" in another, making meta-analysis impossible. After VARC-2, researchers can pool data with greater confidence that endpoints are measured consistently.

In practical terms, the impact is substantial: a 2014 study found that VARC definitions were used in over 90% of TAVI trials published after 2011. The standardisation reduced heterogeneity in outcome reporting, allowing more reliable comparisons between transcatheter and surgical approaches.

**Acknowledged by authors:**

The definitions are based on expert consensus, not empirical data

Some definitions may need further refinement as evidence accumulates

The document does not address all potential complications (e.g., silent cerebral embolism detected by MRI)

Quality of life assessment recommendations are preliminary and not fully validated in TAVI populations

**Critical reader observations:**

Industry representatives participated in the consensus process, creating potential conflicts of interest—definitions that minimise apparent complication rates could favour device manufacturers

No patient representatives were included, so definitions may not reflect what matters most to patients (e.g., a "minor" bleeding event might be terrifying to a patient)

The document is now over a decade old; subsequent evidence has challenged some definitions (e.g., the troponin threshold for peri-procedural MI is now considered too high by some experts)

No formal validation studies were conducted to test whether these definitions predict long-term outcomes

The consensus process lacked transparency about how disagreements were resolved and whether dissenting opinions were recorded

The document focuses on clinical endpoints but does not address patient-reported outcomes beyond a brief mention of quality of life instruments

Definitions for some endpoints (e.g., valve thrombosis) have been substantially revised in subsequent VARC-3 documents (2021)

For someone running their own n=1 experiment (e.g., testing a lifestyle intervention to improve cardiovascular health):

**What to test:**

If you are recovering from aortic valve replacement (surgical or transcatheter), test whether a structured cardiac rehabilitation programme (e.g., 12 weeks of supervised exercise 3×/week) improves functional capacity compared with usual care

Alternatively, test whether a Mediterranean diet intervention (with specific targets: ≥5 servings vegetables/day, ≥3 servings fish/week, olive oil as primary fat) reduces inflammatory markers compared with your current diet

**Minimum meaningful duration:**

For functional capacity: 12 weeks of exercise training, with measurements at baseline, 6 weeks, and 12 weeks

For dietary intervention: 8–12 weeks to see changes in inflammatory markers (hs-CRP, IL-6); 6 months for clinically meaningful changes in blood pressure or lipid profile

For any endpoint defined in VARC-2 (e.g., bleeding, stroke), you cannot meaningfully test these in an n=1 experiment—they are too rare and require large populations

**What to measure (specific metrics):**

Functional capacity: 6-minute walk test (distance in metres), Duke Activity Status Index (0–58.2 points, higher = better)

Quality of life: Kansas City Cardiomyopathy Questionnaire (KCCQ, 0–100, higher = better, minimal clinically important difference = 5–10 points)

Inflammatory markers: high-sensitivity C-reactive protein (hs-CRP, mg/L), interleukin-6 (IL-6, pg/mL)

Blood pressure: seated resting systolic/diastolic (mmHg), measured at same time of day, after 5 minutes rest, average of 3 readings

Heart rate variability: 5-minute recording using a validated device (e.g., Polar H10 chest strap), measuring RMSSD (ms) and SDNN (ms)

**Key confounds to control for:**

Medication changes: any new or discontinued cardiovascular medications (statins, beta-blockers, ACE inhibitors, antiplatelet agents) can dramatically affect outcomes

Activity level: track daily step count using a pedometer or smartphone; changes in physical activity can confound dietary interventions

Sleep quality: poor sleep increases inflammation and blood pressure; track sleep duration and quality using a sleep diary or wearable

Stress: acute stress raises cortisol and blood pressure; use a daily stress rating (0–10 scale) or the Perceived Stress Scale (PSS-10) weekly

Seasonal effects: blood pressure and inflammation vary with season (higher in winter); run your experiment during a single season or control for outdoor temperature

Alcohol and caffeine: both affect blood pressure and heart rate; standardise intake or track daily consumption

Timing of measurements: always measure at the same time of day, ideally in the morning before food, caffeine, or exercise

**What a positive result would look like:**

For functional capacity: increase in 6-minute walk distance of ≥30 metres (minimal clinically important difference in heart failure patients) or increase in KCCQ score of ≥5 points

For inflammatory markers: reduction in hs-CRP of ≥1.0 mg/L (if baseline >2.0 mg/L) or reduction in IL-6 of ≥0.5 pg/mL

For blood pressure: reduction in systolic BP of ≥5 mmHg (clinically meaningful for cardiovascular risk reduction)

For heart rate variability: increase in RMSSD of ≥10 ms (indicating improved parasympathetic tone)

A positive result should be consistent across at least 3 measurements at the end of the intervention period, and should exceed the variability seen during a 2-week baseline period

**Important caveat for n=1 experiments:**

The VARC-2 definitions are designed for large clinical trials with hundreds or thousands of patients. They define rare but serious events (death, stroke, major bleeding) that cannot be meaningfully studied in a single person. For your own experiment, focus on continuous, modifiable outcomes like functional capacity, quality of life, blood pressure, and inflammatory markers. Use the VARC-2 document as a reminder that standardised definitions matter—define your outcomes clearly before you start, and stick to those definitions throughout your experiment.