Coronary microvascular dysfunction

Article written and reviewed by Dr Luca Moretti

Coronary microvascular dysfunction, abbreviated CMD, is the inability of the heart's small vessels to adequately regulate perfusion and resistance. It involves prearterioles, arterioles, and capillaries that are not resolved by conventional coronary angiography and may result from structural or vasomotor abnormalities or their combination; the term pathophysiological abnormality does not by itself identify a syndrome. CMD may be symptomatic, silent, primary, or secondary to other diseases.
CMD is not synonymous with microvascular angina. The latter requires an ischemic presentation attributable to the microcirculation, typically in the absence of an epicardial stenosis capable of explaining the symptoms; CMD may instead be detected in obstructive coronary artery disease, myocardial infarction, cardiomyopathy, aortic stenosis, heart failure, or after revascularization. This clinical boundary prevents overdiagnosis. An abnormal value describes a function; it does not automatically demonstrate the cause of pain.

The microcirculation accounts for most coronary resistance and continuously adapts flow to oxygen consumption. Because no single test measures all its components, diagnosis integrates flow, reserve, resistance, vasoreactivity, epicardial anatomy, and myocardial context; multidomain assessment is more informative than an isolated threshold. Adenosine and acetylcholine interrogate different mechanisms and may produce discordant but valid results.
Clinical interest stems from the frequency of CMD and its associations with ischemia, persistent angina, diastolic dysfunction, heart failure, and adverse prognosis; these associations do not, however, justify considering every low CFR value an autonomous disease or prescribing universal therapy. Contextual causality must be reconstructed. The meaning of the same measurement differs among ANOCA, acute myocardial infarction, cardiomyopathy, and an asymptomatic patient.

Etiology, risk factors, pathogenesis, and pathophysiology

Epicardial arteries primarily conduct blood, while progressively smaller vessels modulate resistance. Prearterioles respond to pressure and endothelial signals, arterioles to local metabolism, and capillaries distribute oxygen and substrates; the unfavorable transmural gradient makes the subendocardium more vulnerable. Coronary regulation combines myogenic, metabolic, endothelial, and neurohumoral control. A defect at any level may produce heterogeneous perfusion without visible narrowing.
During pharmacological hyperemia, a healthy microcirculation markedly reduces resistance and increases flow several-fold. In structural CMD, hyperemia remains limited because of inward arteriolar remodeling, medial thickening, perivascular fibrosis, extravascular compression, or capillary loss, with minimal resistance often elevated. The structural phenotype primarily reduces maximal flow. Hypertension, diabetes, hypertrophy, and aging promote this configuration, but no hemodynamic index is equivalent to histological examination.

Functional CMD may present with reduced reserve despite preserved hyperemic resistance. In many cases, the denominator is excessive resting flow caused by tachycardia, greater cardiac work, autonomic dysregulation, or inappropriately low vascular tone, so the relative margin for increase narrows; the functional phenotype does not imply vessels that are normal in every respect. Absolute flow, heart rate, blood pressure, and the rate-pressure product help interpret reduced CFR.
The endothelium releases nitric oxide, prostacyclin, and hyperpolarizing mediators, counterbalanced by endothelin, thromboxane, and reactive species. Oxidative stress, inflammation, modified lipoproteins, and insulin resistance reduce nitric oxide bioavailability and promote leukocyte adhesion, thrombogenicity, and vasoconstriction. Endothelial dysfunction affects arteries and microvessels. It may precede evident plaque, but an abnormal endothelial response does not necessarily coincide with clinical ischemia or diagnosed spasm.

Microvascular spasm is dynamic vasoconstriction that can be provoked with acetylcholine and is recognized when typical symptoms and ischemic ECG changes occur without epicardial narrowing of 90 percent or greater. It may coexist with reduced adenosine reserve or epicardial spasm, creating overlapping endotypes. Dynamic vasoconstriction differs from fixed remodeling. Autonomic activity, smooth muscle, the endothelium, inflammation, and receptor sensitivity contribute to varying degrees.
Hypertension and ventricular hypertrophy increase intramyocardial compression and oxygen demand, while rarefaction and fibrosis prevent a proportional increase in supply. Diabetes adds glycation, mitochondrial dysfunction, autonomic neuropathy, and capillary loss; obesity and inactivity sustain inflammation and high resting flow. Metabolic determinants act on multiple compartments. Blood-pressure or glycemic control is essential, but an incomplete response does not demonstrate that the initial mechanism was incorrect.

Chronic kidney disease, anemia, sleep apnea, smoking, dyslipidemia, and systemic inflammation alter oxygenation, tone, and microvascular structure. Lupus, rheumatoid arthritis, and other immune-mediated diseases may combine endothelial inflammation, accelerated atherosclerosis, microthrombosis, and iatrogenic effects, requiring association to be distinguished from the dominant causal mechanism. The systemic context guides investigations and treatment. CMD must not become a label that replaces diagnosis of the primary disease.
Hypertrophic cardiomyopathy, amyloidosis, aortic stenosis, and other forms of hypertrophy cause compression, reduced relative capillary density, and high demand; dilated cardiomyopathies and myocarditis may alter the endothelium and metabolic coupling. In heart failure with preserved ejection fraction, inflammation, stiffness, and CMD form a bidirectional network that is not yet fully resolved. Associated heart diseases change prognosis. In these patients, low reserve does not automatically justify a diagnosis of primary microvascular angina.

In obstructive coronary artery disease, reduced distal pressure, embolization of atherothrombotic material, vasoconstriction, and inflammation impair the downstream bed. After PCI or bypass surgery, microemboli, edema, ischemia-reperfusion injury, and vasomotor abnormalities may occur, while after myocardial infarction microvascular obstruction persists despite reopening of the epicardial artery. Secondary CMD modifies recovery and residual symptoms. It must not be interpreted using the same diagnostic objectives as in a person with stable ANOCA.
Age and sex affect vessel size, resting flow, endothelial response, and prevalence of associated conditions. Women are highly represented in ANOCA and INOCA studies, but CMD occurs in both sexes and in numerous settings other than angina without obstruction; pretest probability depends on the population examined. Estimates obtained at tertiary centers or in female cohorts cannot be transferred to the general population without adjustment.

Pathological anatomy, classification, and progression

Resistance vessels have diameters below the resolution of coronary angiography, so CMD is defined mainly through functional surrogates. Biopsy and autopsy studies show medial hypertrophy, luminal narrowing, perivascular fibrosis, deposits, capillary loss, and endothelial abnormalities in selected settings, but findings vary with the cause. The anatomical basis is heterogeneous and rarely sampled in vivo. Elevated resistance supports, but does not by itself demonstrate, remodeling or rarefaction.
Physiological classification separates a structural endotype, characterized by low reserve and high hyperemic resistance, from a functional endotype with low reserve and relatively normal resistance; this distinction helps interpret the relationship between resting and hyperemic flow, but thresholds and measurements do not create perfectly discrete biological categories. Adenosine endotypes form a continuum. Discordant Doppler and thermodilution values require technical review and interpretation, not an arithmetic mean.

A third domain is endothelium-dependent vasomotion, which includes inadequate vasodilation and inducible spasm. Microvascular spasm may occur with normal adenosine indices, while a patient with high resistance may also develop vasoconstriction with acetylcholine; the most useful classification therefore permits overlaps. The vasomotor domain complements reserve. Labeling only one endotype may miss a therapeutically relevant component.
The Camici and Crea classification distinguishes four etiological settings: CMD without obstructive coronary artery disease or myocardial disease, CMD associated with myocardial disease, CMD associated with obstructive coronary artery disease, and iatrogenic CMD. It emphasizes that the same hemodynamic abnormality has a different meaning in cardiomyopathy, after myocardial infarction, or in ANOCA. The four settings are not mutually exclusive. A patient may move from one group to another with atherosclerotic progression or interventions.

In ST-elevation myocardial infarction, microvascular obstruction is defined on CMR as a hypointense area within tissue showing late gadolinium enhancement and represents severe ischemia-reperfusion injury. Intramyocardial hemorrhage, edema, neutrophils, microthrombi, and distal embolization sustain its persistence despite TIMI 3 epicardial flow. Microvascular obstruction is a specific acute phenotype. It must not be confused with chronic CMD diagnosed by CFR in stable patients.
In infiltrative or hypertrophic cardiomyopathies, the arteriolar wall may be thickened and compressed, while capillary density becomes insufficient relative to mass. In aortic stenosis, pressure load, reduced diastolic perfusion, and hypertrophy combine to reduce reserve, which may improve only partially after valve replacement. The capillary-to-mass ratio helps explain ischemia. Prognosis also remains dominated by the structural heart disease and its treatment.

CMD may be regional, diffuse, or heterogeneous. An invasive measurement in the left anterior descending artery samples one territory, while PET and CMR describe perfusion globally or by segment; a global average may conceal regional defects, and a single artery may not represent the entire heart. Spatial distribution affects sensitivity and comparison. The report must state territory, stressor, quality, and presence of diffuse epicardial disease.
Progression is not necessarily linear. Autonomic tone, medications, inflammation, and hemodynamic status may rapidly modify the functional component, while rarefaction and fibrosis change more slowly; after revascularization or valve replacement, some indices improve without normalizing. Microvascular plasticity varies with the mechanism. Repeating a test makes sense only if the result can change a decision and the method and conditions are comparable.

Reduced reserve may precede evident clinical manifestations in diabetes, kidney failure, or cardiomyopathy, but indiscriminate screening of asymptomatic people has not demonstrated an event benefit. An incidental finding therefore calls for quality verification, assessment of the cause, and risk management, not automatic antianginal therapy. Subclinical CMD is a potential marker. It remains to be demonstrated which test-based strategies improve outcomes compared with standard care of the associated disease.

Clinical manifestations

CMD may be asymptomatic and emerge during PET, CMR, or invasive assessment performed for another reason. When it causes ischemia, manifestations include chest pain, pressure, dyspnea, fatigue, and reduced capacity, but no symptom specifically identifies the microcirculation rather than stenosis, epicardial spasm, or noncardiac causes. Absence of symptoms does not exclude a physiological abnormality. At the same time, an abnormal test does not necessarily explain an existing complaint.
In patients with non-obstructive coronary arteries, CMD may underlie ANOCA when symptoms predominate and INOCA when ischemia is documented. When symptoms, ischemia, and the microvascular abnormality are causally consistent, microvascular angina is present; COVADIS criteria classify that syndrome, not every form of CMD. The anginal manifestation is only one clinical phenotype; this distinction avoids conceptual conflation and guides the choice of therapeutic endpoints.

Pain attributable to the microcirculation may occur with exertion, emotional stress, cold, or at rest, vary in duration, and resolve slowly after activity. The response to nitroglycerin is inconsistent because arterioles, perception, and demand contribute differently from epicardial spasm, and it is not a diagnostic test. The symptom profile guides but does not confirm. ECG, troponin, and anatomy remain necessary when the pattern or intensity suggests an acute event.
In heart failure with preserved ejection fraction or hypertrophy, exertional dyspnea and intolerance may jointly reflect elevated filling pressure, chronotropic incompetence, reduced peripheral extraction, and CMD. Attributing all dyspnea to the microcirculation is therefore inappropriate even when reserve is low; multifactorial dyspnea requires complete phenotyping. Echocardiography, natriuretic peptides, and cardiopulmonary exercise testing answer questions different from coronary measurements.

In hypertrophic cardiomyopathy, CMD may contribute to angina, dyspnea, exertional syncope, and fibrosis, but dynamic obstruction, end-diastolic pressure, arrhythmias, and high consumption are competing explanations. In aortic stenosis, similar symptoms also depend on the gradient, hypertrophy, and low perfusion pressure; the dominant syndrome remains structural heart disease. The microvascular finding refines its pathophysiology without replacing established intervention criteria.
After PCI, a successfully treated epicardial artery does not guarantee disappearance of angina. Preexisting CMD, microembolization, vasospasm, diffuse disease, and noncoronary causes may sustain residual or recurrent symptoms, while restenosis and thrombosis must be excluded when indicated. Post-PCI angina requires an orderly diagnosis. Automatically repeating a procedure without evidence of epicardial ischemia may add risk without providing benefit.

In acute myocardial infarction, no-reflow and microvascular obstruction may present with failure of ST-segment resolution, persistent pain, instability, or incomplete ventricular recovery despite opening the artery. The presentation is time-dependent and is assessed with angiography, ECG, and especially CMR in appropriate patients. Reperfusion injury belongs to a different clinical phase. CFR thresholds used in stable ANOCA are not mechanically applied during the event.
In heart transplantation, endothelial and microvascular dysfunction may accompany graft vasculopathy, rejection, or remodeling, while denervation attenuates angina. Silent ischemia, ventricular dysfunction, or reduced capacity may therefore precede typical symptoms and require specialist transplant-center protocols. Cardiac denervation changes the presentation. Monitoring is not replaced by criteria developed for nontransplant patients.

Palpitations, syncope, edema, or pleuritic pain are not specific manifestations of CMD and should trigger investigation for arrhythmias, heart failure, pulmonary embolism, pericarditis, or other causes. Reflux, chest-wall disorders, and panic attacks may also coexist with an abnormal microvascular measurement; the differential diagnosis remains mandatory. A plausible physiological explanation must agree with timing, ischemia, territory, and response, not merely with an out-of-range number.
New prolonged pain, diaphoresis, severe dyspnea, syncope, hypotension, or a neurological deficit requires urgent assessment. Previous CMD does not protect against atherothrombotic myocardial infarction, spontaneous coronary dissection, embolism, Takotsubo syndrome, myocarditis, acute aortic syndrome, or pulmonary thromboembolism; warning signs take precedence over the chronic diagnosis. Comparing the new episode with the usual pattern reduces both underestimation and unnecessary repeated visits.

Investigations and diagnosis

Before the microcirculation is measured, the question is defined: explaining angina without obstruction, assessing a cardiomyopathy, stratifying myocardial infarction, or investigating symptoms after PCI. History, examination, ECG, laboratory tests, and echocardiography identify causes of high demand, ventricular dysfunction, hypertrophy, and valvular heart disease that alter flow and prognosis. The diagnostic question determines the technique and threshold. The same result does not produce the same diagnosis in different populations.
CCTA and coronary angiography describe the epicardial lumen but do not directly visualize arterioles and capillaries. In suspected ANOCA or INOCA, an intermediate stenosis must be assessed physiologically: FFR above 0.80 makes focal ischemia from that lesion unlikely without excluding diffuse disease or CMD. Epicardial anatomy remains an interpretive prerequisite. Non-obstructive plaque and microvascular dysfunction may coexist and both contribute to risk.

Coronary flow reserve is the ratio of flow or velocity during hyperemia to the resting value. CFR below 2.0 is often used with PET and some invasive methods, whereas Doppler and thermodilution in the 2024 ESC guidelines commonly use the threshold below 2.5; integrated CFR reflects the epicardial and microvascular compartments. The method, territory, stressor, and threshold must appear in the report.
CFR may fall because maximal flow is low or because resting flow is high. Heart rate, blood pressure, anemia, pain, caffeine, nicotine, medications, fasting, and inadequate hyperemia affect the ratio; age, sex, and ventricular mass modify its distribution even without a clear pathological threshold. Hemodynamic confounders must be recorded. A value near the limit requires pretest probability and concordant data, not a binary label.

IMR is calculated with a pressure guidewire and thermodilution as the product of distal pressure and mean transit time during hyperemia; a value of 25 or greater is commonly abnormal. Compared with CFR, it depends less on resting flow but is not completely independent of territory, collaterals, stenosis, and technique. Hyperemic IMR supports elevated microvascular resistance. Its isolated prognostic value is less consistent than that of reserve and varies with context.
HMR divides distal pressure by average peak Doppler velocity during hyperemia. ESC 2024 reports a value above 2.5 mmHg·s/cm as abnormal, but studies have used different protocols and limits; the Doppler signal requires expertise and stability. Doppler resistance is not interchangeable with IMR. Both sample one territory and must be corrected for an epicardial lesion when present.

Doppler and thermodilution measure related but different phenomena and show only moderate agreement. Conventional thermodilution estimates flow from bolus transit and may overestimate reserve relative to Doppler, while continuous thermodilution permits absolute flow and resistance measurement but requires specialist equipment and validation. Technical discordance is not eliminated by choosing the preferred number. Curve quality, stability, and reproducibility must be reviewed before clinical interpretation.
PET quantifies myocardial blood flow in mL/min/g at rest and under stress and calculates myocardial flow reserve, commonly abnormal below 2.0. It is the most validated noninvasive method for quantification and prognosis, but the global value includes diffuse atherosclerosis, blood pressure, cardiomyopathy, and resting flow in addition to the microcirculation. Quantitative PET provides robust stratification. Radiation, tracer, availability, motion correction, and software limit access and comparisons among centers.

Stress CMR assesses perfusion, function, edema, and scar without radiation. In WISE, MPRI below 1.84 distinguished CMD with 73 percent sensitivity and 74 percent specificity; quantitative studies have proposed MPR below 2.19 and prognostic thresholds such as MPRI of 1.47 or lower. CMR thresholds are not universal. Scanner, sequence, stress, artifacts, postprocessing, and population prevent a single limit from being transferred to every laboratory.
Transthoracic Doppler usually measures velocity in the left anterior descending artery during adenosine or dipyridamole. CFVR below 2.0 or 2.5 suggests an abnormality according to the protocol; iPOWER identified a prognostic threshold below 2.25 in women with angina and no obstruction. Transthoracic CFVR is repeatable and involves no radiation. The acoustic window, expertise, single territory, and assumption of a constant diameter limit generalizability.

Intracoronary acetylcholine assesses endothelium-dependent vasomotion and propensity to spasm, whereas adenosine assesses predominantly endothelium-independent vasodilatory capacity. Low doses, often between 2 and 20 micrograms, assess the endothelial response; higher doses, up to 100 or 200 micrograms according to the protocol, provoke vasoconstriction. Acetylcholine testing requires continuous monitoring. The dose, artery, time, symptoms, ECG, and angiographic change must be documented.
Epicardial spasm requires reproduction of symptoms, ischemic ECG changes, and constriction of 90 percent or greater. Microvascular spasm has the first two elements without that constriction; a simple reduction in flow, an attenuated endothelial response, or pain alone does not satisfy the complete definition. Vasospastic criteria prevent ambiguous labels. A negative response does not exclude adenosine endotypes, and a positive response may coexist with them.

At experienced laboratories, provocative testing has a favorable safety profile but may induce bradycardia, block, atrial fibrillation, tachyarrhythmias, or persistent spasm. Takahashi's meta-analysis of 12,585 patients estimated major events at 0.5 percent with no deaths; the 2025 Dutch registry reported major events in 0.9 percent with no procedural death, myocardial infarction, or stroke. Expert safety is not the same as harmlessness. Indication, consent, preparation, and rescue drugs remain essential.
A TIMI frame count greater than 25 was included among the COVADIS signs of MVA, but is not a reliable substitute for CFR, IMR, or HMR. Contemporary studies show poor accuracy of coronary slow flow for detecting invasive CMD because injection, heart rate, anatomy, caliber, and epicardial resistance affect frame counts. Slow flow is insufficient to diagnose CMD. It may generate suspicion that requires appropriate physiology, not a definitive conclusion.

There is no single gold standard for every component of CMD. PET and CMR describe global or regional perfusion, Doppler measures velocity, IMR and HMR hyperemic resistance, and acetylcholine vasomotion; biopsy is not routinely indicated, and histology does not automatically coincide with function. Integrated diagnosis combines concordant measurements and context. A robust conclusion states what has been demonstrated, what has been inferred, and what remains an alternative.
ESC 2024 recommends invasive functional testing, class I level B, in persistently symptomatic patients with suspected ANOCA or INOCA, inadequate response to therapy, and impaired quality of life to identify treatable endotypes. PET, CMR, or Doppler of the left anterior descending artery may be considered for reserve, class IIb B. Test selection depends on the clinical consequence; these recommendations are not equivalent to screening every asymptomatic person for CMD.

Treatment and prognosis

CMD has no single therapy because it is not a single syndrome. The first objective is to treat the dominant cause or heart disease, then reduce risk and symptoms according to physiology; in myocardial infarction, reperfusion is managed; in aortic stenosis, the valve disease; in cardiomyopathy, the specific phenotype; and in ANOCA, the coronary endotype. The therapeutic hierarchy begins with context. Normalizing an index is not a validated clinical endpoint in every population.
Smoking cessation, adapted exercise, diet, sleep, and weight management support vascular health and capacity. Blood pressure, lipids, diabetes, kidney disease, anemia, sleep apnea, and inflammation are treated according to their own guidelines, with individual goals and attention to hypotension or hypoglycemia that may worsen perfusion. Global prevention is fundamental even without symptoms. Evidence that one lifestyle measure alone normalizes CMD and reduces events remains indirect.

In patients with angina, reduced reserve, and high resistance, beta-blockers are commonly used to reduce demand, with calcium channel blockers, ACE inhibitors or ARBs, and ranolazine selected according to heart rate, blood pressure, ventricular function, and comorbidities. ESC 2024 considers drugs targeting demand ischemia for symptoms in MVA with reduced flow, class IIa B. Hemodynamic selection limits random trials. There is no definite evidence that these drugs reduce MACE in isolated CMD.
When low CFR primarily results from high resting flow with preserved hyperemic resistance, control of heart rate, demand, and conditions that increase cardiac work is physiologically plausible. A beta-blocker or ivabradine may be considered in selected cases, but bradycardia, hypotension, and chronotropic incompetence may worsen capacity. Resting flow must be interpreted before prescribing. Randomized evidence specific to this endotype remains limited.

In the spastic microvascular component, calcium channel blockers are the cornerstone of symptom therapy; nitrates or nicorandil may be added, especially when epicardial spasm coexists, if available and tolerated. Isolated beta-blockade, particularly with a nonselective agent, may be unfavorable in uncontrolled spasm and requires caution. Vasodilator therapy follows documentation and response. Headache, edema, hypotension, and nitrate tolerance affect titration.
ACE inhibitors and statins may improve some endothelial indices and are appropriate when hypertension, diabetes, kidney disease, ventricular dysfunction, plaque, or indicated risk is present. Small trials, including WISE with quinapril, suggest benefit for CFR and angina in subgroups but do not demonstrate reduced death or myocardial infarction from CMD. Vascular protection retains conventional indications. A drug package is not prescribed automatically solely for low reserve.

Aspirin is not mandatory therapy for CMD without atherosclerosis, a previous event, PCI, or another indication. Potential antithrombotic benefit must be weighed against bleeding, age, and concomitant therapy; functional microvascular dysfunction is not equivalent to unstable plaque or ongoing thrombosis. The antithrombotic decision remains etiological. Revascularization and stenting likewise do not treat the microvascular bed in the absence of a significant epicardial lesion.
In 151 patients, CorMicA showed that disclosing functional test results and linking them to stratified therapy improves angina and quality of life at six months, with persistence at one year. ILIAS ANOCA reported an improved Seattle Angina Questionnaire score with a similar strategy in 2025; the endotype-guided strategy therefore has evidence for symptoms. These studies were not powered for mortality, myocardial infarction, or heart failure and involved symptomatic people.

Pharmacological evidence remains heterogeneous. Ranolazine was neutral overall in a crossover study of 128 participants, with a signal in the subgroup with lower CFR; ChaMP-CMD showed that physiology may identify exercise responses to amlodipine or ranolazine, whereas the PRIZE trial with zibotentan did not improve capacity and produced adverse effects. A subgroup response generates hypotheses, not universal certainty. Different endpoints, durations, and definitions limit direct comparisons.
WARRIOR randomized 2,476 women with suspected ANOCA or INOCA to systematic intensive therapy with a high-intensity statin, an ACE inhibitor or ARB, and aspirin or to usual care. The composite was not reduced, with a hazard ratio of 1.13 and a confidence interval of 0.94-1.37; underenrollment and contamination reduce precision. WARRIOR's neutral result argues against indiscriminate prescription. It does not invalidate statins or renin-angiotensin system blockers when a proven indication exists.

Cardiac rehabilitation, progressive increases in activity, and psychological support may reduce deconditioning and the effect of symptoms. The objective is monitored through episode frequency, rescue-drug use, distance, daily activities, and instruments such as the Seattle Angina Questionnaire, without confusing well-being with physiological normalization. Patient-reported outcomes are clinically important. Symptom improvement does not demonstrate structural regression, just as stable CFR does not imply rehabilitation failure.
Prognosis depends on the population. Meta-analyses of observational studies associate isolated CMD with increased mortality and MACE, while PET shows a continuous relationship between myocardial flow reserve and outcomes beyond anatomical categories; causality, thresholds, and treatment nevertheless remain influenced by comorbidities and diffuse disease. A prognostic gradient is more realistic than a sharp cutoff. An individual estimate requires age, ventricular function, plaque, kidney function, diabetes, and clinical setting.

Follow-up is adapted to the cause, symptoms, and risk: clinical and risk-factor reassessment, adherence, adverse effects, and ECG or imaging when the presentation changes. Systematically repeating CFR, IMR, or PET to demonstrate improvement is not supported by outcome evidence and may expose patients to procedures or radiation without therapeutic consequence. Targeted monitoring prioritizes useful decisions. New manifestations nonetheless require the differential diagnosis to be reopened.

Complications

Repeated ischemia and maldistribution of flow may promote diastolic dysfunction and fibrosis, particularly when hypertrophy, diabetes, or inflammation coexist. The relationship is bidirectional: elevated filling pressures compress the microvessels, and CMD may in turn worsen relaxation and reserve during exertion; the diastolic cycle does not demonstrate a single initial cause. Dyspnea and edema require diagnosis of heart failure using independent criteria.
In the Taqueti study, the combination of CFR below 2 and diastolic dysfunction identified a more than fivefold risk of future hospitalization for HFpEF. The finding is observational and does not prove that therapy targeting CFR prevents heart failure, but supports closer assessment of at-risk patients; transition to HFpEF concerns a subgroup. Blood pressure, obesity, atrial fibrillation, and kidney disease remain fundamental prognostic targets.

After myocardial infarction, microvascular obstruction and intramyocardial hemorrhage are associated with more extensive necrosis, adverse remodeling, heart failure, and arrhythmias despite successful epicardial reperfusion. CMR allows them to be quantified and refines prognosis, while numerous experimental interventions against no-reflow have not produced a universally effective therapy. Persistent no-reflow represents severe injury. Prevention, rapid reperfusion, and post-infarction management follow specific protocols.
In stable coronary artery disease or after PCI, CMD may contribute to residual ischemia, recurrent pain, and repeated procedures. If a microvascular measurement is not integrated with stent assessment, epicardial physiology, and noncardiac causes, the patient risks both unnecessary further revascularization and minimization of a genuinely significant stenosis. The risk of misclassification operates in both directions. A sequential pathway reduces errors without promising to eliminate every symptom.

In patients without obstruction, CMD is associated with major cardiovascular events in cohorts and meta-analyses, but endpoints include different combinations of death, myocardial infarction, stroke, heart failure, and hospitalization for angina. Differences in technique, threshold, and referral explain some variability and preclude a single percentage valid for everyone. MACE risk must be contextualized. It cannot be inferred from pain alone or a marginally abnormal value.
Persistent angina or dyspnea reduces activity, work, sleep, and quality of life and increases urgent visits, tests, and costs; this morbidity is particularly relevant when patients receive conflicting explanations after non-obstructive coronary angiography or when the physiological diagnosis is not linked to a verifiable plan. The healthcare burden is a concrete complication. Precise communication avoids both prognostic alarmism and the inappropriate phrase ‘normal heart.’

Empirical polypharmacy may cause hypotension, bradycardia, edema, headache, fatigue, electrolyte disturbances, kidney injury, or QT prolongation; these effects may mimic worsening heart disease or further reduce exercise tolerance, especially when several agents are started simultaneously. Drug-induced harm requires surveillance. Blood pressure, heart rate, kidney function, potassium, ECG, and interactions guide goal-directed titration.
The tests themselves have undesirable consequences: radiation with PET, contrast and invasiveness with catheterization, gadolinium or claustrophobia with CMR, complications from acetylcholine, and false results near thresholds. The choice must consider informational benefit, procedural risk, and the concrete possibility that the result will change management. Diagnostic harm also includes incorrect labels. Technical quality and transparent reporting are clinical safety measures.

A diagnosis of CMD may obscure new diseases if applied automatically to every episode. Atherosclerotic progression, thrombosis, spontaneous dissection, embolism, myocarditis, Takotsubo syndrome, arrhythmia, valvular heart disease, or pulmonary disease may emerge independently of the previous finding; clinical change requires reassessment. Continuity means using known physiology as context, not as the mandatory explanation for every future symptom.
The principal epistemic complication is turning a marker into a certain cause. CFR, MFR, IMR, and HMR are continuous variables influenced by anatomy, hemodynamics, and method; a threshold facilitates decisions but does not perfectly separate health from disease or demonstrate that correcting the value improves survival. Inferential caution protects the patient. Better diagnosis and therapy arise from convergence of physiology, context, symptoms, and evidence of efficacy.