Diagnosis of ischemic heart disease

Article written and reviewed by Dr Luca Moretti

Diagnosing ischemic heart disease does not coincide with simply looking for a coronary stenosis because it must establish whether the patient has myocardial ischemia, which mechanism produces it, whether irreversible injury exists, and what the immediate and future risks are. The same manifestation may result from acute atherothrombosis, stable epicardial disease, vasospasm, microvascular dysfunction, or a systemic imbalance between oxygen supply and demand; conversely, a plaque or stenosis may be present without explaining the symptom.
The correct pathway therefore integrates history, physical examination, ECG, biomarkers, ventricular function, coronary anatomy, and physiology, selecting examinations according to the clinical question. The objective is not to accumulate tests but to reach an etiological diagnosis that materially changes prevention, pharmacological therapy, revascularization, or follow-up.

The first branch distinguishes a possible acute coronary syndrome, in which time determines the amount of salvageable myocardium, from a stable chronic presentation. Ongoing or worsening pain, hemodynamic instability, ischemic ECG changes, arrhythmias, acute heart failure, or a dynamic increase in troponin requires an urgent pathway; reproducible, stable symptoms instead permit probabilistic estimation and a rational choice between coronary CT angiography and functional imaging.
This distinction is not rigid: chronic disease may become destabilized, and an apparently acute presentation may result from myocarditis, pulmonary embolism, aortic dissection, or Takotsubo syndrome. Diagnosis therefore remains serial and is updated when new data modify the probability of the initial hypotheses.

Etiology, risk factors, pathogenesis, and pathophysiology

Coronary atherosclerosis is the most common substrate and arises from intimal retention of apoB-containing lipoproteins, the inflammatory response, and wall remodeling. Progressive plaque formation may narrow the lumen or remain outwardly remodeled until rupture or erosion occurs; on a thrombogenic surface, platelets and coagulation generate a thrombus capable of reducing or interrupting blood flow.
Diagnosis must distinguish atherosclerotic burden from the culprit lesion: coronary calcium and nonobstructive plaque document disease and risk, whereas an intermediate stenosis often requires functional assessment before being considered the cause of ischemia.

Ischemia occurs when oxygen supply does not meet demand. Coronary blood flow depends on perfusion pressure, epicardial and microvascular resistance, and diastolic duration; demand increases with heart rate, contractility, and wall tension. Tachycardia, hypotension, anemia, and hypoxemia may therefore precipitate ischemia even without new thrombosis, especially in the presence of preexisting coronary artery disease.
The 2026 Fifth Universal Definition of Myocardial Infarction requires separation of myocardial injury, identified by troponin, from infarction, in which acute injury must be attributed to ischemia through clinical, electrocardiographic, anatomical, or imaging findings.

An epicardial stenosis produces a pressure loss dependent on geometry, length, irregularity, and blood flow; angiographic diameter alone does not reliably quantify this effect. FFR measures the pressure ratio during hyperemia; iFR and other nonhyperemic indices are measured without adenosine, but not all assess the diastolic phase: some use the entire cardiac cycle, allowing a lesion to be linked to a physiological consequence.
In diffuse disease, the gradient is distributed along the vessel and a focal target may be absent; physiological pullback distinguishes this pattern from a localized pressure drop and reduces the risk of implanting a stent in a circulation that will remain globally limited.

Microvascular dysfunction includes reduced vasodilation, increased resistance, rarefaction, and arteriolar remodeling. It may present with nonobstructive epicardial coronary arteries or coexist with stenosis, diabetes, hypertrophy, cardiomyopathy, or previous infarction; reduced flow reserve does not identify the mechanism by itself because it depends on both resting blood flow and hyperemic capacity.
Coronary spasm is dynamic epicardial or microvascular vasoconstriction, often episodic, that may not occur during an examination performed outside the attack and, in selected cases, requires controlled intracoronary provocation.

Age, sex, symptoms, smoking, blood pressure, lipids, diabetes, kidney function, family history, and vascular disease modify clinical probability but do not replace diagnosis. Pretest probability serves to avoid two mirror-image errors: underestimating plausible disease and subjecting patients at very low risk to a cascade of false-positive results.
Probability must be updated after each examination by means of its likelihood ratio; a result is not intrinsically positive or negative but acquires meaning in the context of technical quality and prior probability.

Clinical manifestations and initial assessment

The history defines site, quality, duration, radiation, and precipitating and relieving factors. Retrosternal tightness provoked by exertion or stress and relieved by rest increases the likelihood of ischemia, but dyspnea, nausea, sweating, weakness, or epigastric pain may represent equivalents, especially in older adults, women, diabetes, and kidney disease.
The term atypical pain must not become synonymous with benignity: pleuritic, positional, or reproducible features reduce the probability of a coronary origin without excluding it when the overall context remains high risk.

The temporal sequence is decisive. Sudden pain that is maximal at onset suggests aortic dissection or embolism; a recent crescendo, episodes at rest, or a clear reduction in threshold requires assessment for acute coronary syndrome. In chronic disease, the provoking activity, functional class, frequency, duration, response to nitrate, and change over time are documented.
Syncope, palpitations, and paroxysmal dyspnea may indicate an ischemic arrhythmia or heart failure; the absence of pain does not exclude silent ischemia, particularly in diabetes and patients with autonomic neuropathy.

Physical examination begins with consciousness, perfusion, blood pressure, heart rate, oxygen saturation, and signs of distress. Blood pressure differences, pulse deficits, or a new diastolic murmur require consideration of aortic dissection; hypotension, cold skin, oliguria, and altered mental status indicate shock. Cardiac and pulmonary auscultation looks for a third heart sound, new murmurs, crackles, and effusion.
A new systolic murmur after infarction may reflect acute mitral regurgitation or a septal defect, whereas jugular venous distention with relatively clear lung fields points toward right ventricular involvement.

Assessment includes immediately dangerous alternative diagnoses: aortic dissection, pulmonary embolism, pneumothorax, tamponade, esophageal rupture, and sepsis. Reflux, musculoskeletal pain, radiculopathy, and anxiety are common but are considered after risk assessment and not as an automatic explanation for a normal initial ECG.
Clinical probability arises from integration rather than a single descriptor; even a noncoronary diagnosis may coexist with significant atherosclerosis and modify preventive management.

Medications, substances, and comorbidities must be reconstructed. Cocaine and amphetamines promote spasm and thrombosis; triptans and vasoconstrictors may precipitate episodes in susceptible individuals; anemia, thyrotoxicosis, infection, and tachyarrhythmias increase demand. Previous PCI or CABG modifies the anatomy and the choice of test.
Documentation of stents, grafts, ventricular function, and previous images avoids repetition and makes it possible to recognize a true change from baseline.

Investigations and diagnosis

In an acute presentation, a 12-lead ECG is acquired rapidly and repeated during symptoms or when the clinical picture evolves. Persistent ST-segment elevation, dynamic ST-segment depression, new T-wave inversions, or posterior and right-sided changes indicate the territory and urgency; a normal or nondiagnostic tracing does not exclude ischemia.
Posterior and right-sided leads increase sensitivity in selected settings, while comparison with previous ECGs helps distinguish new changes from bundle branch blocks, hypertrophy, or chronic repolarization abnormalities.

High-sensitivity cardiac troponin is interpreted using an algorithm validated for the specific assay and serial samples. A value above the reference limit indicates injury, whereas a dynamic change suggests acuity; diagnosing infarction nevertheless requires evidence of ischemia. Sex, time from onset, kidney function, and chronic values influence interpretation.
The 2026 Fifth Universal Definition recommends sex-specific 99th percentile limits and clinical classification as primary, secondary, or procedure-related infarction, replacing the automatic use of the biomarker as a synonym for atherothrombosis.

Echocardiography assesses global and regional function, the right ventricle, valves, and complications. A new segmental abnormality supports ischemia but may be absent in small infarctions and may be preexisting; echocardiographic contrast improves endocardial definition when the window is poor.
During instability, bedside echocardiography rapidly identifies mechanical shock, tamponade, and congestion; in chronic disease, it provides the prognostic measurement of ejection fraction and guides selection of the next test.

In a stable patient with a low-to-intermediate probability, coronary CT angiography is particularly effective for excluding obstructive disease and documents nonobstructive plaque, anatomy, and anomalies. Extensive calcium, a high heart rate, small stents, and artifacts may overestimate stenosis; kidney function, allergy, and radiation enter into the choice.
The calcium score modifies probability and risk, but a score of zero does not completely exclude noncalcified plaque or disease in symptomatic young people; CCTA must lead to a decision, not an isolated description.

Stress echocardiography, CMR, SPECT, and PET look for functional consequences. Echocardiography identifies inducible contractile abnormalities; CMR combines perfusion, function, and scar without radiation; PET quantifies absolute blood flow and reserve; SPECT is widely available but uses relative distribution and may underestimate balanced ischemia.
The choice depends on ECG, exercise capacity, obesity, kidney function, devices, expertise, and the clinical question; adding different tests without a reason increases discordant results and incidental findings.

Exercise ECG testing retains selective utility when the tracing is interpretable and the patient can achieve an adequate workload, providing information on symptoms, capacity, blood pressure, and arrhythmias. It has lower accuracy than imaging and does not reliably localize microvascular disease or balanced ischemia.
A result is judged together with the achieved workload, heart rate, blood pressure response, and recovery; early termination for noncardiac reasons makes the test nondiagnostic, not negative.

Invasive coronary angiography is indicated without delay when acute risk is very high or ischemia persists, and in chronic presentations when refractory symptoms, high risk, or inconclusive tests make revascularization plausible. It shows the lumen, not the wall or causality directly; intermediate lesions often require FFR or iFR.
IVUS and OCT define dimensions, calcium, plaque, dissection, and the stent result; intracoronary imaging is used when the information changes the diagnosis or procedure, avoiding indiscriminate instrumentation.

In the absence of an explanatory obstruction, the pathway does not end. CFR and microvascular resistance assess vasodilator capacity and the microvascular compartment; acetylcholine provokes vasomotion and makes it possible to document epicardial or microvascular spasm in an experienced laboratory. The final diagnosis distinguishes ANOCA, INOCA, microvascular angina, and vasospastic angina.
Early CMR in presentations with troponin elevation and nonobstructive coronary arteries identifies infarction, myocarditis, or Takotsubo syndrome; MINOCA remains a working diagnosis until the etiology is defined.

According to the 2024 ESC and 2021 ACC/AHA guidelines, diagnosis requires a pathway consistent with the clinical context:

• immediately recognize acute presentations and time-dependent alternative diagnoses;
• estimate clinical probability and risk before selecting the test;
• document injury, ischemia, anatomy, and function using complementary methods, without treating them as synonyms;
• physiologically assess intermediate stenoses when the result guides revascularization;
• investigate microvascular disease and vasospasm when the epicardial coronary arteries do not explain the symptoms;
• formulate a final etiological diagnosis and state the residual uncertainty.

The differential diagnosis of injury with troponin elevation includes myocarditis, Takotsubo syndrome, heart failure, pulmonary embolism, sepsis, kidney failure, and tachyarrhythmia. The distribution of late enhancement on CMR distinguishes an ischemic subendocardial or transmural scar from nonischemic patterns, while edema and function establish the timing of the process.
The diagnosis of ischemic heart disease must finally define the target organs: ventricular function, mitral regurgitation, arrhythmias, kidney function, the brain, and systemic vascular disease modify prognosis and therapy.

Diagnostic management, stratification, and prognosis

The appropriate test is the one capable of changing a decision. In acute syndromes, the diagnostic pathway proceeds simultaneously with monitoring, antiplatelet therapy, and reperfusion when indicated; delayed certainty is not awaited if the ECG and clinical picture require immediate intervention. In chronic disease, the initial strategy may be anatomical or functional according to probability and patient characteristics.
A sequential strategy avoids duplication: normal coronary CT angiography often ends the search for obstruction, whereas an intermediate stenosis leads to FFR-CT, stress imaging, or invasive physiology according to consequences and availability.

Prognostic risk derives from ventricular function, disease extent, left main or multivessel involvement, plaque burden, ischemia, scar, kidney function, and diabetes. A test may have prognostic value without demonstrating that correcting the finding improves survival; this distinction is essential when proposing a procedure.
In chronic presentations, randomized studies have shown that an initial invasive strategy does not universally reduce death or infarction compared with medical therapy, although it provides greater improvement in angina in symptomatic patients; anatomies and subgroups with a prognostic indication require specific assessment.

The report must describe technical quality, residual probability, and limitations. Expressions such as "negative test" without the workload, achieved heart rate, or artifacts may produce false reassurance; "significant stenosis" based solely on visual assessment may lead to unnecessary revascularization.
Shared decision-making links the result to distinct objectives: preventing death or infarction, reducing angina, clarifying a diagnosis, or avoiding further examinations. Informational value depends on the ability to address one of these objectives.

Prevention begins when atherosclerosis, even nonobstructive, is documented, with control of lipids, smoking, blood pressure, diabetes, activity, and weight according to risk. An incidental finding must not be ignored, but neither should it automatically be regarded as the cause of pain; prevention and diagnosis of the symptom may proceed on two different planes.
Follow-up reassesses new symptoms, adherence, and function, avoiding routine serial tests in stable patients when the result would not change management.

The prognosis after a negative assessment depends on the method and time horizon. CCTA showing no plaque is associated with very low risk, whereas normal stress imaging reduces short-term event risk but does not exclude future progression; reduced flow reserve may retain prognostic significance even without stenosis.
Every result therefore has a period of clinical validity conditioned by new symptoms, events, and factor control, not an automatic expiration that is the same for everyone.

Multimodality integration and diagnostic pathways in different clinical scenarios

In patients with chronic symptoms, assessment begins with the risk factor-weighted clinical likelihood, which integrates age, sex, symptom characteristics, and risk factors instead of automatically applying the older tables based solely on typical pain. Contemporary populations undergoing assessment have a lower prevalence of obstructive coronary artery disease than historical cohorts; overestimating probability therefore exposes patients to unnecessary examinations and procedures, whereas ignoring modifiers such as diabetes, peripheral artery disease, kidney failure, or coronary calcium may produce the opposite error. The estimate serves to select the method with the highest yield, not to formulate the definitive diagnosis.
A very low probability may justify no further testing for obstructive disease after acute causes have been excluded and alternative explanations assessed. At low or moderate probabilities, coronary CT angiography is often preferable because of its high rule-out power and ability to document nonobstructive plaque; at higher probabilities, in patients with known coronary artery disease, or when quantification of ischemia and viability is relevant, functional imaging may provide an answer more directly connected to the decision.

Baseline investigations are not a formality. A complete blood count identifies anemia, blood glucose and glycated hemoglobin define glucose metabolism, and a lipid profile and kidney function guide prevention and contrast safety, while TSH, electrolytes, and other tests are ordered according to the context. The resting ECG may show Q waves, repolarization abnormalities, blocks, or arrhythmias, but it may be normal despite significant disease. Echocardiography describes ejection fraction, wall motion, hypertrophy, valves, and pressures, recognizing disorders that mimic or worsen angina.
A chest radiograph does not diagnose coronary artery disease but is useful when dyspnea, heart failure, aortic disease, or lung disease enters the differential diagnosis. Ambulatory ECG monitoring is reserved for palpitations, syncope, or suspected vasospastic ischemia with transient changes, avoiding interpretation of the absence of changes during an asymptomatic period as exclusion of disease.

Coronary CT angiography must be acquired with a controlled heart rate, appropriate synchronization, and reconstructions capable of reducing artifacts. The report states quality, dominance, vessel origin, plaque distribution, degree of stenosis, and high-risk findings; CAD-RADS classification may standardize communication and the subsequent pathway but does not replace clinical judgment. Dense calcifications produce blooming and may overestimate narrowing, while motion, obesity, and irregular rhythm reduce segment readability. An apparently severe stenosis in a nonevaluable segment must not automatically become an indication for PCI.
Calcium quantification using the Agatston score describes calcific atherosclerotic burden and adds prognostic information but does not always localize the culprit lesion and does not detect the noncalcified component. In symptomatic patients, the calcium score may refine probability or help interpret a test, whereas indiscriminate repeated screening does not demonstrate the evolution of every plaque. Anatomical information must translate into preventive intensity and, when necessary, functional verification.

CT-derived FFR applies fluid-dynamic models to anatomical images and may estimate the hemodynamic effect of intermediate stenoses without new invasive access. Quality nevertheless depends on segmentation and acquisition; artifacts that limit CCTA also compromise the calculation. Stress CT perfusion is another way to combine anatomy and function, at the cost of additional contrast and radiation. These techniques are most useful when they resolve a specific discordance and reduce angiograms without treatable lesions.
The general principle is that a 50-70% stenosis does not have a physiological significance predictable from the visual percentage. Length, caliber, serial disease, perfused myocardial mass, and the microcirculation determine the pressure drop and ability to increase blood flow. Anatomy therefore describes the substrate, while physiology establishes whether and how it limits perfusion.

Stress echocardiography looks for new wall-motion abnormalities provoked by exercise, dobutamine, or a vasodilator and makes it possible to correlate symptoms, blood pressure, and function. Exercise is preferable when the patient can perform it because it reproduces the physiological stimulus and provides functional capacity; a poor acoustic window requires echocardiographic contrast or an alternative method. Sensitivity may be reduced in mild single-vessel disease, while bundle branch block, pacing, and cardiomyopathies complicate interpretation of wall motion.
SPECT and PET assess perfusion distribution. SPECT compares territories relatively, and a balanced global reduction may appear less evident, although transient dilation, a fall in ejection fraction, and other clues suggest high risk. PET offers better resolution and quantifies resting blood flow, hyperemic blood flow, and reserve; a global reduction in reserve may result from diffuse epicardial disease, microvascular dysfunction, or both and requires integration with anatomy and hemodynamic conditions.

Stress magnetic resonance imaging combines first-pass perfusion, function, edema, and late gadolinium enhancement. An inducible defect without corresponding enhancement supports ischemia in viable myocardium; subendocardial or transmural enhancement in a coronary distribution documents an ischemic scar, whereas mid-wall or subepicardial patterns point toward nonischemic causes. The finding makes it possible to distinguish pain due to inducible ischemia, previous silent infarction, myocarditis, and cardiomyopathy, provided that quality, artifacts, and contraindications are considered.
Scar transmurality provides information on the likelihood of segmental recovery after revascularization, but viability does not automatically equate to a clinical benefit from the procedure. The decision depends on symptoms, anatomy, global function, operative risk, and the amount of myocardium at risk; turning a tissue finding into an isolated indication confuses a prognostic association with a treatment effect.

During invasive coronary angiography, the two-dimensional projection may underestimate eccentric lesions and does not describe the wall. FFR during hyperemia assesses the ratio of distal to aortic pressure, whereas iFR and other nonhyperemic indices are measured at rest using different algorithms; values close to the thresholds must be interpreted with signal quality, drift, guide-catheter pressure damping, serial lesions, and the clinical picture. FFR/iFR discordance is not necessarily an error because the two indices respond differently to blood flow, the microcirculation, and the disease pattern.
Slow pullback distinguishes focal pressure loss, potentially correctable with a stent, from a gradual drop due to diffuse disease. After PCI, physiology may identify underexpansion, residual lesions, or diffuse pressure loss, while IVUS and OCT define the anatomical mechanism. Integration reduces both treatment of noncausal stenoses and a technically suboptimal result from an indicated procedure.

In patients with angina and nonobstructive coronary arteries, diagnosis must investigate a vasomotor endotype. Coronary reserve assesses the overall capacity to increase blood flow; IMR and other indices estimate microvascular resistance; acetylcholine identifies epicardial vasoconstriction when it produces marked narrowing with symptoms and ischemic changes, or microvascular spasm when symptoms and ECG changes occur without sufficient epicardial constriction. Reduced vasodilation and hyperreactivity may coexist, accounting for mixed phenotypes and different therapeutic responses.
Invasive coronary function testing requires appropriate withdrawal of medications, standardized protocols, and a laboratory prepared to treat bradycardia, prolonged spasm, or arrhythmias. Its value lies not only in assigning a name but in replacing the generic label of noncardiac pain with a mechanistic diagnosis that guides calcium channel blockers, nitrates, risk-factor control, and treatment of microvascular dysfunction.

In acute coronary syndrome, probability and urgency are established simultaneously. Persistent ST-segment elevation compatible with occlusion, or an equivalent presentation with instability, requires a reperfusion strategy without awaiting biomarkers; in suspected NSTE-ACS, high-sensitivity troponin, serial ECGs, and validated scores distinguish early rule-out, observation, and an invasive strategy. The troponin delta must be interpreted according to the assay and time from onset: a low initial value very early may require a subsequent sample, whereas chronically elevated concentrations in kidney disease do not nullify the significance of a dynamic change.
Coronary angiography may show thrombosis, dissection, embolism, or no obstruction. In MINOCA presentations, OCT or IVUS may identify rupture, erosion, or a subtle dissection, while early CMR distinguishes infarction from myocarditis and Takotsubo syndrome. The pathway ends only when the mechanism has been defined sufficiently to guide therapy and prognosis.

Some populations require adaptations. In women, a lower prevalence of epicardial obstruction for the same symptoms coexists with a higher frequency of microvascular dysfunction and vasospasm; using nonobstructive coronary angiography as proof of the absence of disease perpetuates diagnostic delays. In diabetes and older adults, dyspnea, asthenia, confusion, or heart failure may replace pain, while neuropathy and reduced exercise capacity limit the reliability of symptoms and exercise tests.
In chronic kidney disease, atherosclerosis, calcification, baseline troponin, and contrast-related risk increase; the choice must balance yield, iodine volume, and the possibility of a genuine therapeutic decision. After CABG, CCTA may assess graft patency well but is less accurate for calcified native vessels, and functional imaging helps localize the territory when anatomy is complex. None of these conditions justifies the same automatic sequence for everyone.

The quality of the pathway is also measured by clinical-diagnostic concordance. A positive examination in a patient with a very low probability may be false, whereas a normal test with inadequate stress does not sufficiently reduce a high probability. Collegial review of images, comparison among territories, and verification of raw data are preferable to indiscriminate repetition of examinations. If anatomy and function are discordant, it is necessary to ask whether the lesion is diffuse, whether the microcirculation is impaired, whether the image contains artifacts, or whether the symptom has another cause.
The final report should state the most likely diagnosis, alternatives excluded, degree of certainty, risk, and recommended consequence. A pathway ending with "normal coronary arteries" is incomplete if documented angina persists; similarly, nonobstructive plaque requires prevention but must not be presented as the definite cause of every pain. This separation among presence of disease, symptom mechanism, and future risk is the core of modern diagnosis.

Complications and diagnostic errors

The most serious complication is failure to diagnose an acute coronary syndrome, promoted by nonclassic symptoms, a normal initial ECG, sampling too early, or premature attribution to anxiety or reflux. Women, older adults, patients with diabetes, and patients with kidney disease are particularly exposed to less recognizable presentations.
Serial pathways with ECG and troponin, reassessment of pain, and return instructions reduce risk, but no algorithm replaces judgment when instability or clinical change develops.

A false-positive result may lead to coronary angiography, contrast exposure, radiation, bleeding, and revascularization without benefit. Very low probability, artifacts, bundle branch block, hypertrophy, and inadequately performed tests reduce specificity; repeating methods without resolving the discordance amplifies the cascade.
A multimodality review identifies whether the discordance depends on anatomy, physiology, quality, or an alternative mechanism.

A false-negative result may derive from balanced ischemia, inability to achieve adequate stress, episodic vasospasm, microvascular dysfunction, or a lesion not present at the time of examination. A reassuring result must not nullify a high clinical probability; in these cases, a method based on a different principle is selected or an invasive approach is taken when appropriate.
Recognition of INOCA has reduced the error of regarding patients as normal when coronary angiography shows only the absence of obstructive stenoses.

Iodinated contrast may cause reactions and kidney injury, while radiation and incidental findings accompany CCTA, SPECT, and invasive procedures. Pharmacological stress may cause hypotension, bronchospasm, or arrhythmias; exercise testing may precipitate significant ischemia, although risk is low in appropriate settings.
Safety depends on selection, preparation, access to resuscitation, and timely termination; consent provides information about risk and diagnostic alternatives.

An imprecise diagnosis produces psychological, insurance-related, and therapeutic consequences. Labeling nonischemic injury as infarction may result in unjustified antithrombotic therapy and restrictions; failure to recognize a secondary infarction or silent scar prevents appropriate prevention and follow-up.
Final documentation must separate the finding, syndrome, and cause, state what has been demonstrated and what remains probable, and provide for diagnostic reassessment when new data become available.