Cirrhotic cardiomyopathy

Article written and reviewed by Dr Luca Moretti

Cirrhotic cardiomyopathy is a syndrome of myocardial dysfunction that develops in the setting of cirrhosis and is not explained by a pre-existing primary heart disease. At rest, the heart may appear hyperdynamic, with a normal or elevated ejection fraction, whereas physiologic or therapeutic stress reveals an inability to adequately increase cardiac output, heart rate and relaxation. This apparent contradiction between high output and reduced cardiac reserve constitutes the clinical core of the disease.

The modern definition does not indiscriminately include every cardiovascular abnormality observed in cirrhosis. Tachycardia, increased stroke volume, volume-related atrial dilation, prolonged corrected QT interval or a modest rise in natriuretic peptides may reflect hepatic physiology, medications, age or a comorbidity. To attribute dysfunction to cirrhosis, a consistent phenotype must be demonstrated and alternatives must be investigated systematically. The diagnosis of exclusion is therefore a positive component of the reasoning process, not a residual formula.

The condition often remains subclinical because systemic vasodilation reduces the load against which the ventricle ejects blood. Infection, gastrointestinal bleeding, exercise, uncompensated paracentesis, creation of a portosystemic shunt or liver transplantation abruptly change loading conditions and metabolic demands. At those times, an apparently stable balance may turn into congestion, hypoperfusion or decompensation. The hemodynamic latency explains why a reassuring baseline examination does not complete the assessment in high-risk patients.

The concept is useful mainly when it guides verifiable decisions: correctly interpreting the echocardiogram, estimating risk before TIPS or transplantation, avoiding volume overload, and distinguishing potentially reversible injury from independent heart disease. However, there is no single reference test or validated specific pharmacologic therapy. A multidisciplinary assessment remains necessary because the liver, heart, kidney and pulmonary circulation participate in the same clinical picture without being interchangeable.

Definition, nosology and boundaries of the syndrome

Cirrhotic cardiomyopathy is defined as systolic and/or diastolic dysfunction in a patient with cirrhosis, in the absence of heart disease capable of explaining it. The term does not indicate a single morphology, because it may present with preserved ejection fraction, abnormal strain, delayed relaxation, elevated filling pressures or an attenuated chronotropic response. The phenotypic heterogeneity prevents it from being reduced to a classic dilated cardiomyopathy.

The presence of cirrhosis is necessary, but it does not automatically establish causality for any finding. Hypertension, diabetes, obesity, kidney disease, coronary artery disease, valvular disease and atrial fibrillation become more common with age and with metabolically associated fatty liver disease. These conditions can produce the same heart-failure phenotype with preserved ejection fraction. A cirrhotic causal relationship becomes credible when anatomy, function, clinical history and the hemodynamic profile converge after reasonable exclusion of alternatives.

It should not be confused with congestive hepatopathy or cardiac cirrhosis, in which disease of the right heart, valves or pericardium causes venous congestion and liver injury. In cirrhotic cardiomyopathy, the predominant causal direction runs from liver disease and its systemic physiology toward the myocardium. The direction of causality changes the diagnosis, prognosis and indication for combined transplantation.

High-output heart failure is not synonymous with this diagnosis either. Cirrhosis can produce high cardiac output, chamber dilation and congestive symptoms through vasodilation and sodium and water retention without proving intrinsic cardiomyocyte injury. Conversely, myocardial dysfunction may be present before clinical heart failure develops. Distinguishing hemodynamic phenotype from cardiomyopathy prevents mechanisms that overlap in only some patients from being used as synonyms.

Hyperdynamic circulation and cardiac reserve

Portal hypertension promotes splanchnic vasodilation through nitric oxide, carbon monoxide, prostanoids and other mediators, while portosystemic shunts expand the functional vascular bed. Systemic vascular resistance falls and effective arterial blood volume is perceived as inadequate despite expansion of extracellular volume. The result is increased heart rate, stroke volume and cardiac output at rest. The hyperdynamic circulation is therefore a compensatory response to an abnormal distribution of blood.

The baroreflex, sympathetic nervous system, renin-angiotensin-aldosterone system and vasopressin support blood pressure and perfusion through extrasplanchnic vasoconstriction and sodium and water retention. Over time, preload increases while effective afterload remains low. The ventricle may increase in mass and volume and maintain an apparently vigorous ejection fraction. This load dissociation makes ejection fraction an incomplete indicator of intrinsic contractility.

When exercise, sepsis or bleeding requires a further increase in output, the response no longer has the same reserve. Adrenergic desensitization limits the increase in heart rate and contractility, while a stiff ventricle accepts additional volume only by increasing filling pressures. The result may be low effective perfusion and congestion at the same time. The inotropic reserve therefore cannot be inferred from the high cardiac output measured under baseline conditions.

Stress assessment is also complex. Nonselective beta-blockers, anemia, sarcopenia, ascites, deconditioning and vasodilation alter performance and test endpoints; dobutamine may fail to achieve target heart rate and exercise may stop for noncardiac reasons. An abnormal result describes reduced global reserve but does not by itself prove a cirrhotic mechanism. The physiology under stress must be interpreted together with diastolic function, strain, blood pressure and ongoing therapy.

Cellular, autonomic and inflammatory mechanisms

Experimental models show reduced density or coupling of beta-adrenergic receptors, altered G proteins and lower cyclic AMP production in response to stimulation. Increased membrane cholesterol reduces sarcolemmal fluidity and interferes with receptors and channels. These findings explain part of the attenuated inotropic response, but they come mainly from animals with bile duct ligation or toxic exposure. Translation to humans remains plausible but has not been definitively quantified.

Calcium homeostasis is altered at the level of L-type channels, the sarcoplasmic reticulum and the sodium-calcium exchanger. Reduced systolic availability limits force, while inefficient reuptake prolongs relaxation. Changes in potassium currents and action potential contribute to delayed repolarization. The calcium-dependent dysfunction thus provides a biological link between reduced contractility, diastolic dysfunction and QT prolongation.

Intestinal bacterial translocation and endotoxemia activate monocytes, endothelial cells and pathways such as NF-kB. Tumor necrosis factor alpha, interleukins and inducible nitric oxide synthase increase nitric oxide and cGMP in the myocardium, with cardiodepressant and nitrosative effects. Reactive oxygen species, mitochondrial dysfunction and apoptosis may perpetuate injury. The immunometabolic signal becomes particularly relevant during infection and acute-on-chronic liver failure.

Endocannabinoids activating the CB1 receptor, circulating bile acids, carbon monoxide and endothelins have been implicated in modulation of contractility. Their relative importance varies with the model, etiology and disease stage, and none currently constitutes a validated diagnostic biomarker or approved therapeutic target. The plurality of mediators suggests a pathogenic network rather than a single responsible toxin.

Chronic neurohormonal activation, volume overload and inflammation promote hypertrophy, collagen deposition and titin modifications, increasing passive stiffness. Autonomic neuropathy and attenuation of the baroreflex also reduce the ability to adapt heart rate and vascular tone. The interstitial fibrosis observed with mapping and extracellular volume in human CMR studies supports structural involvement, but is not specific to cirrhosis and requires comparison with age and comorbidities.

Systolic, diastolic, chronotropic and electrical phenotypes

Overt systolic dysfunction at rest is less common than the term cardiomyopathy might suggest because low afterload preserves ejection fraction. Global longitudinal strain may detect less effective shortening before ejection fraction falls, but it too depends on preload and afterload. In markedly vasodilated cirrhosis it may even become more negative. The subclinical systolic function therefore requires standardized acquisition and interpretation of loading conditions.

The diastolic phenotype includes delayed relaxation, reduced mitral annular e' velocity, increased filling pressures and atrial remodeling. In early stages, the patient compensates by increasing the atrial contribution; with progression, volume loading or tachycardia precipitates congestion and pulmonary edema. Age and hypertrophy, however, produce the same findings. Advanced diastolic dysfunction is more specific than a single abnormal relaxation index.

Chronotropic incompetence describes the inability to adequately increase heart rate during exercise or pharmacologic stimulation. Beta-adrenergic desensitization, dysautonomia and beta-blocker therapy may all contribute and cannot always be separated. Electromechanical coupling may also be delayed. The chronotropic response has pathophysiologic significance, but the Consortium placed it among areas requiring validation rather than among the core diagnostic criteria.

QTc prolongation is the best-known electrical abnormality and may affect a large proportion of patients with advanced disease. It depends on repolarization, autonomic tone, electrolytes, medications and the correction method; Bazett's formula often overestimates QTc during tachycardia, whereas Fridericia is more stable. The QTc in cirrhosis signals vulnerability and warrants pharmacologic caution, but it does not establish cardiomyopathy and does not uniformly predict torsades de pointes.

The right ventricle is exposed to increased flow, filling pressure and possible pulmonary vascular disease. Right ventricular dysfunction may accompany the syndrome, but portopulmonary hypertension, hepatopulmonary syndrome, pulmonary embolism and tricuspid valve disease require separate categories. Fractional area change, TAPSE, right ventricular strain and estimated pressures must be interpreted in context. Biventricular involvement does not justify labeling incompletely defined pulmonary physiology as cirrhotic.

Clinical presentation and factors that unmask the disease

Dyspnea, fatigue, edema and reduced functional capacity are nonspecific because ascites, anemia, sarcopenia, hepatic hydrothorax, encephalopathy and pulmonary disease produce similar symptoms. Orthopnea, crackles, jugular venous distention and response to diuretics suggest cardiac congestion, but their value also changes with hypoalbuminemia and extracellular volume overload. The phenotyping of congestion requires examination, imaging and longitudinal evolution, not symptom description alone.

Sepsis increases vasodilation and cardiodepressant mediators while further reducing effective arterial blood volume. Transient dysfunction during spontaneous bacterial peritonitis may represent decompensation of cirrhotic reserve, septic cardiomyopathy, or both. Variceal bleeding also combines hypovolemia, catecholamines, anemia and resuscitation. Acute decompensation should not be used to retrospectively establish chronic cardiomyopathy without reassessment after stabilization.

Renal failure may result from vasodilation and neurohormonal activation, but an inability to augment cardiac output and elevated venous pressures further impair perfusion and filtration. In prospective cohorts, more advanced diastolic dysfunction has been associated with hepatorenal syndrome and mortality. The association does not prove that correcting an echocardiographic index prevents the event. The cardiorenal connection should prompt more refined volume assessment, not automatic causal attribution.

Surgery, anesthesia and transplantation expose the patient to vasodilatory induction, bleeding, transfusions, clamping and abrupt changes in venous return. Graft reperfusion may cause hypotension, bradycardia, electrolyte abnormalities and myocardial depression. In the following hours, normalization of vascular resistance increases afterload and makes previously masked contractile dysfunction evident. Perioperative stress explains why risk cannot be summarized by preoperative ejection fraction.

From the Montreal criteria to the Consortium

The 2005 World Congress of Gastroenterology consensus proposed, for the systolic component, an attenuated contractile response to exercise, volume loading or pharmacologic stimulation, or an ejection fraction below 55%. For the diastolic component, it listed an age-corrected E/A ratio below 1, deceleration time above 200 ms, or isovolumic relaxation time above 80 ms. The Montreal criteria created a common language but predated tissue Doppler and strain imaging.

Supportive criteria included electrophysiologic abnormalities, chronotropic incompetence, electromechanical uncoupling, prolonged QTc, left atrial enlargement, increased myocardial mass and elevated BNP, proBNP or troponin I. These findings were not equivalent and had numerous confounders. E/A, deceleration time and IVRT change with volume, heart rate and age and have a nonlinear relationship with severity. Their preload dependence contributed to very high prevalence estimates.

The Cirrhotic Cardiomyopathy Consortium formulated proposed criteria in 2019, published in 2020 by Izzy and colleagues. Systolic dysfunction is present if left ventricular ejection fraction is 50% or less or if the absolute GLS value is below 18%. Because GLS is reported as a negative number, -14% is less favorable than -20%. The GLS threshold should always specify the absolute value, software and image quality.

Advanced diastolic dysfunction requires at least three of the four findings listed in the CCC table: septal e' velocity below 7 cm/s, E/e' ratio of 15 or greater, indexed left atrial volume above 34 mL/m², and peak tricuspid regurgitation velocity above 2.8 m/s. The latter criterion applies in the absence of primary or portopulmonary hypertension. The three-criterion rule favors specificity and does not equate mildly delayed relaxation with cardiomyopathy.

Each parameter retains substantial limitations: e' decreases with age and annular calcification, E/e' is an imperfect estimate of filling pressures in cirrhosis, LAVI is affected by chronic volume loading and atrial fibrillation, while tricuspid velocity increases with high flow or pulmonary disease. GLS varies with loading conditions, vendor and frame rate. The criteria are an expert proposal, not a histologic gold standard. Their validation limitations require reporting the underlying data and context in addition to the final label.

The 2025 ASE recommendations updated diastolic assessment in the general cardiology population, but they do not constitute a new disease-specific definition of cirrhotic cardiomyopathy. Likewise, chronotropic response, ECG, biomarkers, chamber measurements, CMR and extracellular volume remain informative elements or research areas according to the CCC. The diagnostic hierarchy therefore distinguishes core criteria, supportive data and tools used to exclude other diseases.

Epidemiology and natural history

There is no universal prevalence. Studies applying the 2005 criteria often classified more than half of patients, whereas the use of tissue Doppler and CCC criteria produces estimates ranging from single digits to about half of patients depending on selection, age, liver disease stage and the GLS definition. Cohorts of transplant candidates do not represent ambulatory compensated cirrhosis. The criteria-dependent prevalence is a methodological fact, not a biological contradiction.

In a transplant cohort analyzed with both definitions, about 77% met the original criteria and 30% the revised criteria, with limited agreement. Other studies found lower rates when GLS was normal and advanced diastolic dysfunction was uncommon, or higher rates when strain and population selection identified greater cardiometabolic risk. The clinical denominator should accompany every percentage to avoid combining screening findings with established disease.

Severity of cirrhosis, ascites and vasoconstrictor activation are often associated with greater remodeling, but not all studies show linear progression. GLS may appear more preserved or hypercontractile in markedly vasodilated stages, while E/e' and LAVI also reflect interventions on volume status. The individual trajectory is more informative than a cross-sectional comparison among patients with different loading conditions.

The natural history includes subclinical stability, development of HFpEF, systolic dysfunction under stress, events after TIPS and peri-transplant decompensation. Etiologic treatment of liver disease may reduce the inflammatory stimulus and portal hypertension, but regression of myocardial injury has not been defined for every cause. Variable reversibility depends on the functional component, fibrosis and competing heart diseases accumulated over time.

Multimodal diagnostic pathway

The evaluation begins with the etiology and stage of cirrhosis, episodes of decompensation, medications, alcohol use, blood pressure, rhythm, functional capacity and signs of congestion. Coronary events, hypertension, diabetes, family history, valvular disease, anemia, infection and renal function should be documented. Weight and volume status change rapidly with diuretics or paracentesis, so the timing of the examination relative to treatment is essential. The loading condition is part of the clinical report.

ECG defines rhythm, conduction, voltages, ischemia and QTc; electrolytes and medications must be documented before attributing repolarization abnormalities to cirrhosis. BNP or NT-proBNP increase with wall stress, volume, age and renal failure, while high-sensitivity troponin may indicate chronic or acute injury. No serum threshold alone diagnoses the syndrome. Cardiac biomarkers are most useful as a baseline, trend and signal for further investigation.

Comprehensive echocardiography measures volumes, mass, biplane ejection fraction, right ventricular function, valves, left atrium and estimated pulmonary pressure. Transmitral and tissue Doppler should include E, A, septal and lateral e', E/e', LAVI and tricuspid velocity; strain requires nonforeshortened views and an adequate frame rate. Echocardiographic standardization enables longitudinal comparison and reduces false changes caused by technique.

Cardiac magnetic resonance is indicated when the acoustic window is poor, volume measurements are discordant, or tissue characterization is needed to exclude myocarditis, infiltration, iron overload or ischemic scar. T1, T2, extracellular volume and LGE may show diffuse edema and fibrosis, but no validated CCC cutoffs exist. Gadolinium requires caution in severe renal dysfunction. Tissue characterization increases etiologic specificity without turning an elevated ECV into exclusive proof of cirrhotic origin.

Cardiopulmonary exercise testing, stress echocardiography or dobutamine CMR can assess reserve in selected cases, particularly when a procedural decision depends on the ability to increase cardiac output. Right-heart catheterization clarifies pressures, output and resistance when echocardiography and clinical findings are discordant or portopulmonary hypertension is suspected. CT coronary angiography or invasive coronary angiography instead addresses the ischemic question. Test selection should begin with the decision that the result could change.

Differential diagnosis and etiologic attribution

Coronary artery disease may be asymptomatic in sedentary individuals and coexist with preserved global function; regional abnormalities, subendocardial scar or risk factors require a dedicated ischemic evaluation. Hypertension, diabetes, obesity and kidney disease independently produce hypertrophy and diastolic dysfunction. Cardiometabolic heart disease is particularly relevant in cirrhosis associated with metabolic dysfunction and should not be absorbed into the cirrhotic label.

In alcohol-related cirrhosis, ethanol exposure can directly cause alcoholic cardiomyopathy with dilation and systolic dysfunction. The presence of cirrhosis does not allow the injury to be automatically attributed to the cirrhotic mechanism, and the two entities may coexist. Cumulative dose, chronology, phenotype, abstinence and recovery guide attribution. Separation from alcohol-related injury is essential for prognosis, counseling and follow-up as well.

Myocarditis, amyloidosis, hemochromatosis, genetic cardiomyopathies, tachycardia-induced cardiomyopathy and drug toxicity require specific clues. CMR, ferritin with transferrin saturation, electrophoresis/immunofixation, rhythm monitoring or genetic testing are used only when justified by the pretest probability. Endocrine cardiomyopathy can also mimic high-output physiology or remodeling. A targeted etiologic workup avoids indiscriminate panels and diagnoses based on incomplete exclusion.

Portopulmonary hypertension and hepatopulmonary syndrome cause dyspnea through different mechanisms and have specific implications for transplantation. Constrictive pericarditis, tricuspid regurgitation and congenital heart disease can cause hepatic congestion and mimic the reverse causal direction. Sepsis, severe anemia and arteriovenous fistulas also produce high output. The hemodynamic differential diagnosis requires distinguishing vascular resistance, filling pressures, gas exchange and ventricular function.

TIPS, acute decompensation and cardiac risk

TIPS rapidly diverts portal blood into the systemic circulation, increasing venous return, central blood volume, right atrial pressure and cardiac output. A heart with impaired relaxation or reduced right ventricular reserve may respond with a marked rise in pressures and heart failure. This effect does not make the procedure harmful in general, but turns latent physiology into a predictable clinical problem. The post-TIPS load justifies cardiac selection proportionate to risk.

Before the procedure, clinicians assess a history of heart failure, dyspnea, arrhythmias, valvular disease and pulmonary hypertension, with ECG and comprehensive echocardiography. BNP or NT-proBNP may add to risk assessment, but renal function and volume status limit the usefulness of fixed thresholds. Severe stage C or D heart failure, untreated severe valvular disease, and moderate or severe pulmonary hypertension despite optimization are contraindications to elective TIPS in North American recommendations. The pre-TIPS assessment therefore integrates right ventricular function, pressures and, when necessary, catheterization rather than stopping at ejection fraction.

The 2025 EASL guidelines and North American recommendations emphasize cardiopulmonary assessment, but no single score replaces clinical judgment. Historical studies linked diastolic dysfunction to poorer outcomes after shunt placement; modern criteria and stents, however, make it difficult to transfer every estimate directly. A controlled initial diameter and hemodynamic reassessment may reduce abrupt overload in appropriate cases. The procedural strategy should be agreed upon by the hepatologist, interventional radiologist, cardiologist and anesthesiologist.

After TIPS, weight gain, orthopnea, hypoxemia, edema, persistent ascites and worsening renal function require differentiation among congestion, encephalopathy, infection and progression of liver disease. Lung ultrasound, echocardiography, serial biomarkers and response to treatment are more informative than an isolated value. Heart failure is treated by balancing diuresis and perfusion; in refractory cases, shunt reduction or occlusion may be necessary. Early monitoring is particularly important in patients with baseline abnormalities.

Liver transplantation and the perioperative period

Transplantation eliminates portal hypertension and liver failure but imposes one of the most intense hemodynamic stresses. The 2026 AASLD/AST guidelines recommend ECG and comprehensive transthoracic echocardiography in the initial evaluation, while strain and tissue Doppler help detect subclinical dysfunction. The 2026 ILTS/Liver Intensive Care Group consensus requires a general cardiovascular assessment for every candidate. Pre-transplant screening is intended to build a plan, not to generate automatic exclusions.

Coronary artery disease, portopulmonary hypertension, valvular disease and overt heart failure must be evaluated separately because they alter candidacy and strategy. Stress tests have reduced sensitivity in the setting of vasodilation and chronotropic incompetence, so the ischemic pathway is adapted to risk and available resources. A positive CCC criterion warrants attention but does not replace coronary anatomy, right ventricular function and invasive assessment when indicated. Integrated stratification avoids attributing every postoperative event to cirrhotic cardiomyopathy.

During surgery, continuous monitoring, calcium and potassium management, vasopressor selection, transfusion control and recognition of reperfusion syndrome are relevant. The ventricle may fail when afterload rises or venous return increases; transesophageal echocardiography and advanced hemodynamic monitoring are selected according to risk and expertise. Perioperative management aims to avoid both hypoperfusion and overload, conditions that may coexist during the same phase.

After transplantation, pulmonary edema, arrhythmias, Takotsubo syndrome, ischemia or new systolic dysfunction may occur. Some patients recover as the circulation normalizes, whereas others retain diastolic dysfunction or remodeling, particularly when fibrosis and cardiometabolic factors were already present. Observational studies have associated the revised CCC criteria with more post-transplant cardiovascular events. Post-transplant reversibility is therefore a possible objective, not an immediate guarantee.

Treatment, follow-up and prognosis

No approved drug has been shown to specifically modify the natural history of cirrhotic cardiomyopathy. The first treatment is etiologic: eliminate the cause of liver disease when possible, prevent decompensation, treat infections and bleeding, and promptly evaluate transplantation in candidates. Control of cirrhosis reduces the stresses that unmask cardiac dysfunction but does not replace cardiologic treatment when true heart failure is present.

Volume management requires small adjustments guided by blood pressure, weight, urine output, sodium, creatinine, pulmonary congestion and perfusion. Diuretics and mineralocorticoid antagonists are often prescribed for ascites, not because they have demonstrated an effect on the cirrhotic myocardium. Large-volume paracentesis and albumin follow hepatology indications. Functional euvolemia is a dynamic target because peripheral edema and low effective arterial blood volume may coexist.

Nonselective beta-blockers reduce portal pressure and prevent decompensation or bleeding in appropriate populations, but they are not specific therapy for cardiomyopathy. Dose and continuation should be reassessed with persistent hypotension, hypoperfusion, acute kidney injury or severe infection. ACE inhibitors and ARBs may worsen blood pressure and renal function in decompensated cirrhosis with ascites. Hemodynamic tolerability takes precedence over automatic application of standard heart-failure regimens.

When documented HFrEF is present, guideline-recommended heart-failure therapies are considered with cautious titration and monitoring of liver and kidney function, potassium and blood pressure; evidence specific to cirrhotic cardiomyopathy is sparse. Atrial fibrillation, thromboembolism and device therapy follow individualized cardiology indications, balancing bleeding and thrombotic risk. Correcting potassium, magnesium and calcium and avoiding combinations that prolong the QT interval reduces preventable risks. Adapted standard therapy is more rigorous than a dedicated empirical treatment.

The CCC proposed echocardiography with tissue Doppler and strain every six months in wait-listed candidates and, when any pre-transplant dysfunction is present, follow-up at 6, 12 and 24 months after surgery. These recommendations are partly consensus-based and should be adapted to disease course, procedure, symptoms and findings. A stable patient with compensated cirrhosis does not automatically require the same interval. Personalized follow-up should have a clinical question and an action threshold.

Prognosis depends on liver disease severity, renal function, congestion, the criteria used and competing heart diseases. In transplant cohorts, the revised criteria have been associated with nearly twice the risk of cardiac events, while reduced septal e' showed particular prognostic value; other studies have not confirmed every association. Probabilistic prognosis cannot be converted into individual destiny or a universal procedural threshold.

The concluding report should describe the phenotype, measurement quality, volume status, alternatives excluded and the scenario that could unmask limited reserve. A robust diagnosis should not merely state cirrhotic cardiomyopathy, but should indicate whether systolic, diastolic, right-sided or chronotropic dysfunction predominates and how this information changes TIPS, transplantation or therapy. Clinical precision consists of transforming an imperfectly defined syndrome into proportionate, verifiable and updatable decisions.