What are the predictor, control, and outcome variables when correlating Cardiac Magnetic Resonance Imaging (Cardiac MRI) with Cardiac Catheterization in patients with suspected or known Coronary Artery Disease (CAD)?

Research Variable Framework for Cardiac MRI vs. Cardiac Catheterization Correlation in CAD

Outcome Variables (Primary Endpoints)

The primary outcome variables should focus on diagnostic accuracy metrics and prognostic indicators that directly impact patient mortality and morbidity. 1

Diagnostic Performance Metrics

• Sensitivity and specificity of cardiac MRI compared to invasive catheterization as the reference standard for detecting hemodynamically significant coronary stenosis (≥50% or ≥70% luminal narrowing) 1
• Positive and negative predictive values for identifying obstructive CAD 1
• Area under the receiver operating characteristic curve (AUC) for overall diagnostic accuracy 1

Prognostic Outcome Variables

• Major adverse cardiac events (MACE) including cardiac death, myocardial infarction, and need for revascularization 1
• All-cause mortality stratified by presence and extent of myocardial damage detected on MRI 1
• Left ventricular ejection fraction (LVEF) and end-systolic volume index (ESVI) as functional outcomes 1

Anatomical and Functional Concordance

• Degree of coronary stenosis (percentage luminal narrowing) on both modalities 1
• Presence and transmural extent of myocardial infarction detected by late gadolinium enhancement on MRI versus wall motion abnormalities on catheterization 1, 2
• Myocardial viability assessment comparing dobutamine stress MRI with catheterization-based functional assessment 1, 2

Predictor Variables (Independent Variables)

Predictor variables represent the cardiac MRI findings being tested against catheterization results. 1

MRI-Derived Anatomical Predictors

• Coronary MRA findings: presence, location, and severity of coronary artery stenosis on magnetic resonance angiography 1
• Late gadolinium enhancement (LGE) pattern: presence, location, and transmural extent of myocardial scar (measured as percentage of LV wall thickness) 1, 2
• Microvascular obstruction detected on first-pass perfusion or LGE imaging 1, 2

MRI-Derived Functional Predictors

• Stress perfusion defects: presence, extent, and severity of inducible ischemia during vasodilator (adenosine/regadenoson) or dobutamine stress 1, 3
• Wall motion abnormalities: regional wall motion score index at rest and with stress 1
• Left ventricular volumes and ejection fraction measured by cine MRI 1
• Myocardial strain parameters if advanced imaging is performed 3

Catheterization-Derived Reference Standards

• Quantitative coronary angiography (QCA) measurements of stenosis severity 1
• Fractional flow reserve (FFR) measurements defining hemodynamically significant stenosis (FFR <0.8) 1
• Intravascular ultrasound (IVUS) findings if performed 1

Control Variables (Confounders to Adjust For)

Control variables must be carefully selected to isolate the true correlation between MRI and catheterization findings. 1

Patient Demographics and Clinical Characteristics

• Age and sex as these affect diagnostic accuracy of both modalities 1
• Body mass index (BMI) and body habitus, which impact MRI image quality 1
• Pre-test probability of CAD using validated risk scores (Framingham, PROCAM, or ESC scores) 4, 5
• Presence of symptoms: chest pain characteristics, dyspnea, or asymptomatic status 1

Cardiovascular Risk Factors

• Traditional risk factors: hypertension, diabetes mellitus, hyperlipidemia, smoking status, and family history 4, 5
• Prior cardiovascular events: previous myocardial infarction, revascularization procedures (PCI or CABG) 1
• Renal function (eGFR/creatinine), as this affects contrast administration and may influence test selection 2

Technical and Image Quality Factors

• Heart rate and rhythm during MRI acquisition, as irregular or high heart rates reduce diagnostic accuracy 1
• Coronary artery calcium score (Agatston score), as heavy calcification significantly reduces specificity of both anatomical imaging modalities 1, 6, 4
• Image quality scores for both MRI and catheterization studies (assessable vs. non-assessable) 1
• Time interval between MRI and catheterization examinations 1

Medication Use

• Beta-blocker use at time of imaging 1
• Antiplatelet and statin therapy status 7
• Vasodilator or inotropic agents used during stress testing 1

Critical Methodological Considerations

Reference Standard Selection

Invasive coronary angiography with FFR should be the primary reference standard rather than angiography alone, as anatomical stenosis severity does not always correlate with hemodynamic significance 1. The ACR Appropriateness Criteria emphasize that FFR-CT shows superior specificity (65-82%) compared to anatomical assessment alone (32-46% specificity) when correlating with invasive FFR 1.

Blinding and Bias Prevention

• Readers of MRI studies must be blinded to catheterization results and vice versa to prevent interpretation bias 1
• Standardized reporting templates should be used for consistency and reliable data collection 1

Common Pitfalls to Avoid

• Do not include patients who are not candidates for revascularization in viability studies, as results will not change management 2
• Avoid performing MRI without gadolinium contrast when assessing for myocardial infarction or viability, as late gadolinium enhancement is essential for distinguishing viable from non-viable myocardium 1, 2
• Recognize that absence of late gadolinium enhancement does not exclude acute ischemia without infarction 2
• Account for the 10% threshold: patients only benefit prognostically from revascularization if the ischemic myocardial area exceeds 10% of LV mass 4