Classification of cardiomyopathies

Article written and reviewed by Dr Luca Moretti

Classifying a cardiomyopathy means transforming findings from anatomy, function, tissue, electrical activity, family history and genetics into a clinical description capable of guiding decisions. The result does not coincide with assigning a name on the basis of echocardiography, because hypertrophy, dilation or scar are phenotypes shared by different processes. A useful classification must make it possible to communicate what the heart shows, separate what is known from what remains uncertain, and indicate which levels of characterization still need to be completed. Its value is therefore measured by its ability to guide etiologic diagnosis, prognosis, therapy and family screening.

The contemporary starting point is the morphofunctional phenotype. This choice does not give morphology biological superiority, but recognizes that imaging, electrocardiography and functional data are available before the cause has been clarified. Once the phenotype is identified, the pathway proceeds toward genetic or acquired etiology, extracardiac involvement, disease phase and risk profile. The final diagnosis thus becomes layered: the same patient may be described, for example, as having a non-dilated left ventricular phenotype, with subepicardial scar, a pathogenic DSP variant, ventricular arrhythmias and familial involvement.

Phenotype and etiology do not have a one-to-one relationship. A sarcomeric defect may produce hypertrophy, restriction or, more rarely, systolic dysfunction; a desmosomal variant may predominantly involve the right ventricle, left ventricle or both; amyloidosis, Fabry disease and glycogen disorders can mimic hypertrophic cardiomyopathy. Conversely, the dilated phenotype may result from TTN, LMNA or many other genes, but also from myocarditis, tachyarrhythmia, alcohol, pregnancy and cardiotoxic agents. Classification must preserve this complexity without becoming an unmanageable list of causes.

Categories are not impermeable biological compartments. Patients may show features of several phenotypes at the same time or move from one to another during the disease course. Scar may precede dilation, hypertrophy may evolve toward thinning and dysfunction, and an initially right-sided disease may become biventricular. Assignment to a category should therefore be accompanied by description of overlapping traits and awareness that the trajectory has clinical meaning that is not fully summarized by the current snapshot.

Classification should not be confused with diagnostic criteria either. The HCM category describes the phenotype of increased wall thickness not explained by loading conditions, whereas thresholds, family findings and supporting elements establish when that phenotype is diagnosable. The ARVC category identifies arrhythmogenic right ventricular disease, but recognition uses a set of structural, tissue, electrocardiographic, arrhythmic and familial criteria. Classification, diagnostic criteria and risk stratification answer different questions and cannot be used as synonyms.

Historical evolution of nosologic systems

Early classifications reflected a cardiology in which pathology and hemodynamics preceded advanced imaging and genetics. The 1995 WHO/ISFC report consolidated dilated, hypertrophic, restrictive and arrhythmogenic right ventricular cardiomyopathy, leaving an unclassified category for conditions that did not fit the major models. Inclusion of ARVC formally recognized a disease in which arrhythmias and right ventricular involvement were not merely variants of dilated cardiomyopathy. The system provided robust clinical terminology for its time, but could not represent the growing molecular diversity and early forms without overt morphology.

The 2006 AHA statement attempted to bring classification closer to biology by distinguishing primary cardiomyopathies, in which the heart is the predominantly involved organ, from secondary cardiomyopathies, in which myocardial injury is part of a systemic disease. Primary forms were further organized into genetic, mixed and acquired groups, and some electrical diseases without evident structural abnormality were also included among genetic forms. This approach made cause explicit, but created difficulties when a protein expressed in multiple organs produced a phenotype clinically dominated by the heart, or when an apparently acquired process acted on an inherited predisposition.

The 2008 ESC position took a different route: beginning with recognizable morphological and functional phenotypes and subdividing them into familial or genetic and non-familial or non-genetic forms. HCM, DCM, ARVC, RCM and unclassified forms constituted the operative categories, while channelopathies were separated because, by definition, they lacked a demonstrated morphological cardiomyopathy. The advantage was applicability to a patient who still lacked an etiologic diagnosis; the limitation was the risk of making categories appear static when genetics and magnetic resonance were showing them to overlap. The comparison with the AHA was not a terminological dispute, but the expression of two different entry points into the problem.

In subsequent years, widespread sequencing demonstrated marked pleiotropy: the same variant or gene can be associated with multiple phenotypes. At the same time, magnetic resonance made non-ischemic scars recognizable before dilation, while electrophysiology documented electrical phases preceding mechanical abnormalities. The old familial/non-familial dichotomy proved insufficient because a negative family history does not exclude a de novo variant, incomplete penetrance or a small family. The distinction between genetic and acquired also became less sharp in gene-environment interaction models.

The MOGE(S) nomenclature, proposed in 2013, sought to represent this multidimensionality through coding analogous to oncology. It allows morphology and function, involved organs, inheritance, etiology and functional status to be reported without forcing all information into a single category. This precision is particularly useful for registries, research and complex families, but the complete string may be less immediate in daily practice. The system nevertheless demonstrates a fundamental principle: a cardiomyopathy diagnosis consists of several axes that should remain distinct.

The 2023 ESC guidelines therefore updated the phenotypic approach by introducing non-dilated left ventricular cardiomyopathy and repositioning hypertrabeculation as a morphological trait. The revision addresses a concrete clinical problem: many patients with non-ischemic scar, arrhythmias or hypokinesia without dilation were not adequately described by DCM or by unclassified categories. The new system does not claim to be definitive, but provides a bridge between the initial finding and etiologic characterization. Its usefulness depends on correct application of definitions and willingness to revise them as the phenotype evolves.

The AHA distinction between primary and secondary became particularly fragile in diseases with variable multiorgan expression. A DES variant may cause dominant cardiomyopathy and mild myopathy in one person, while skeletal muscle may be the main problem in a relative; Fabry disease may initially present as isolated hypertrophy before nephropathy appears. Deciding whether the heart is "predominantly" involved depends on age and time of observation. The cardiac phenotype provides a more stable description, provided it is completed by systemic involvement.

Including channelopathies among AHA primary cardiomyopathies emphasized genetic and electrical relatedness, but broadened the term beyond the presence of myocardial disease. It was later recognized that genes traditionally attributed to channels, such as SCN5A, can produce conduction disease, arrhythmias and dilation, while structural genes may present with cardiac arrest before morphology becomes abnormal. The boundary between occult cardiomyopathy and channelopathy cannot be resolved by the gene list. It requires evidence that a myocardial process exists or is plausible beyond the ion-current abnormality.

"Unclassified" categories had the merit of accommodating new observations, but tended to become heterogeneous repositories. Noncompaction, endocardial fibroelastosis and takotsubo were sometimes grouped together despite different biology, age and reversibility. A residual category is unavoidable as knowledge evolves, but it should be periodically emptied through more precise definitions. Introduction of NDLVC and morphological traits represents exactly this kind of revision.

Architecture of the ESC 2023 classification

The ESC system follows a clinical sequence. First, a presentation is identified, which may consist of symptoms, arrhythmia, electrocardiographic abnormality, an imaging finding, an acute event or family screening; the phenotype is then characterized by echocardiography, magnetic resonance and electrical assessment. The third phase searches for etiology through history, laboratory testing, genetics, specific imaging and, when indicated, histology. Finally, cause and phenotype are translated into risk estimation, therapy and a surveillance plan for the patient and family.

The general definition requires that myocardial abnormalities are not explained exclusively by coronary artery disease, hypertension, valvular disease or congenital heart disease sufficient to produce them. The word "exclusively" is important because these conditions can coexist with cardiomyopathy and modify its expression. Severe aortic stenosis may explain concentric hypertrophy, but not necessarily a familial subepicardial scar; coronary artery disease may cause regional dysfunction, but does not always explain diffuse damage outside ischemic territories. Reasoning should assess proportionality and distribution rather than applying mechanical exclusions.

The phenotype is defined by structural and functional data, but the electrocardiogram and arrhythmias modify its interpretation. Left-sided scar with ventricular tachycardia may carry greater risk than more evident dysfunction that is electrically stable; early atrioventricular block may point toward LMNA or sarcoidosis before echocardiography becomes diagnostic. Classification does not grant imaging a monopoly on diagnosis. On the contrary, it requires an integrated reading in which disproportionate electrical abnormalities become clues to disease and cause.

Etiologic characterization distinguishes genetic from non-genetic causes without necessarily considering them alternatives. The presence of a pathogenic variant may explain the substrate, while myocarditis, exercise, alcohol or pregnancy modulate its penetrance and severity. A non-genetic cause should be supported by exposure, temporality, plausibility and, when possible, response to removal, rather than merely being associated with the patient. The idiopathic category remains acceptable only after a proportionate investigation and should be revisited when new techniques or family information emerge.

Family assessment is part of the classification architecture because it changes pre-test probability and the meaning of borderline findings. A wall thickness of 13 or 14 mm, a limited scar or mild hypokinesia carries different weight in the presence of an affected relative or causal variant. Conversely, apparent clustering may result from shared exposures, and small families may conceal dominant inheritance. Pedigree, clinical screening and molecular segregation should be interpreted together, without turning absence of a known family history into proof of non-heredity.

The system is designed to be updated longitudinally. A carrier may begin as genotype-positive and phenotype-negative, later develop a non-dilated scar and finally dysfunction with dilation; each phase requires a different description, while the etiology remains the same. Recording the trajectory prevents progression from being mistaken for a new unrelated disease. Classification thus becomes a dynamic structure of the clinical record rather than a code assigned once and for all.

The category describes the observed state and not necessarily the entire disease. An LMNA patient with block, arrhythmias and still-normal function may be in a genetic-electrical phase that precedes DCM or NDLVC; forcing that patient immediately into a structural phenotype creates false precision. It is more appropriate to document pre-phenotypic findings and plan surveillance. Classification becomes diagnostic when criteria are met, while in specific genotypes risk may already require attention before that point.

Phenotype assignment requires multimodality imaging when a single method does not visualize the decisive region. Echocardiography defines hemodynamics and valves, magnetic resonance measures the apex and right ventricle and characterizes scar, and computed tomography addresses coronary or pericardial questions. The objective is not to repeat every test in everyone, but to resolve discordances that change category. An apex that is not visualized cannot be classified as normal, and a poorly measured right ventricle cannot establish ARVC.

In children, dimensions and wall thicknesses must be normalized for growth and adult thresholds cannot be used literally. Metabolic phenotypes may change rapidly, and the same sarcomeric variant may become expressed during puberty after initially normal assessments. Classification should also include syndrome, development and neuromuscular involvement because cardiac morphology may be less discriminating. A common language between pediatric and adult cardiology is essential to maintain continuity during transition.

Hypertrophic cardiomyopathy and dilated cardiomyopathy

Hypertrophic cardiomyopathy is defined by increased left ventricular wall thickness, with possible right ventricular involvement, not explained exclusively by loading conditions. In adults, a maximum wall thickness of at least 15 mm supports diagnosis in the appropriate context, whereas values of 13 or 14 mm may be diagnostic when a causal variant, an affected relative or other convincing elements coexist. In children, measurements must be related to body surface area, age and growth using z-scores. The threshold does not replace assessment of distribution, image quality and alternative diagnoses.

The HCM phenotype includes asymmetric septal, apical, concentric, focal and midventricular geometries. Left ventricular outflow tract obstruction is a hemodynamic property that may occur at rest or only with provocation and is defined by a peak gradient of at least 30 mmHg; a value of at least 50 mmHg becomes relevant when considering septal reduction in an appropriately symptomatic patient. Presence or absence of obstruction does not change the underlying category, but identifies subgroups with different mechanisms and therapies. Progression to an ejection fraction below 50% also remains a phase of HCM rather than a new idiopathic dilated cardiomyopathy.

Hypertrophy should be distinguished from athletic adaptation, hypertension and aortic stenosis, and from diseases that mimic the sarcomeric form. Amyloidosis, Fabry disease, Danon disease, Pompe disease, PRKAG2 syndrome and mitochondrial diseases can produce thickened walls but show different ECG, tissue and extracardiac patterns. In phenotypic classification they may be described as HCM when referring to morphology, whereas the etiologic diagnosis should replace vague formulas such as "phenocopy" when the cause is known. This distinction prevents a drug or procedure studied for sarcomeric HCM from being applied indiscriminately.

Dilated cardiomyopathy requires dilation of the left ventricle or both ventricles and global or regional systolic dysfunction not explained exclusively by abnormal loading or coronary artery disease. Volumes should be indexed and compared with appropriate references for sex, age and body surface area, because an absolutely large cavity may be physiological in a large athlete and pathological in a small individual. Dysfunction is not defined by ejection fraction alone when imaging or loading conditions make the measurement unreliable. Morphology, strain, regional wall motion and functional regurgitation complete the picture.

DCM is not synonymous with every form of systolic failure. A large infarction with remodeling, severe valvular overload and congenital heart disease can produce dilation and low ejection fraction without belonging to the category if they adequately explain the phenotype. Conversely, moderate coronary artery disease may coexist with genetic cardiomyopathy when scar distribution and dysfunction do not follow vascular anatomy. The definition should therefore be applied through proportionate causal attribution, not by the simple presence or absence of a competing diagnosis.

Causes of DCM include a broad genetic architecture and numerous acquired factors. TTN, LMNA, FLNC, DSP, RBM20, BAG3, PLN, DES and other genes can produce the same end-stage appearance through different electrical risks and trajectories; myocarditis, alcohol, drugs, tachyarrhythmias and pregnancy may be causes or triggers. Phenotypic classification keeps these presentations together to describe the heart, but management requires separating them again at the etiologic level. Calling an uninvestigated familial form "idiopathic" is not a neutral classification, but an incomplete diagnosis.

Diagnosis of HCM requires that increased wall thickness is real and not caused by adjacent structures included in the measurement; infiltration or deposition may instead produce a hypertrophic phenotype that must be qualified etiologically. The word hypertrophic remains conventional even when the increase results from deposition, because the first level describes appearance; at the etiologic level it should be replaced by a specific formulation. This dual use can create ambiguity unless made explicit. "HCM phenotype due to ATTR amyloidosis" communicates morphology and cause better than either "HCM" or "restrictive" alone.

HCM may show hyperdynamic, normal or reduced systolic function, and ejection fraction does not define the category. A small cavity may eject a high percentage but a modest stroke volume, while strain can be abnormal despite a normal fraction. When ejection fraction falls below 50%, the change indicates systolic progression and modifies therapy and prognosis. Remembering the prior hypertrophic phenotype prevents this phase from being incorrectly reclassified as newly developed DCM.

In DCM, dilation involves the left ventricle and must be accompanied by global or regional systolic dysfunction; right ventricular dilation or dysfunction may be associated but is not required for diagnosis. A large cavity with normal function and reserve in an athlete does not fulfill the pathological concept, whereas a low ejection fraction in a non-dilated heart may fall within NDLVC even in the presence of non-ischemic scar; in the absence of scar, NDLVC includes isolated global hypokinesia. Indexing and reference values reduce the effect of body size. Acute loading conditions and rhythm should be stabilized when possible before defining a permanent phenotype.

NDLVC, ARVC and restrictive cardiomyopathy

Non-dilated left ventricular cardiomyopathy, abbreviated NDLVC, includes non-ischemic scar or fatty replacement of the left ventricle, irrespective of the presence of global or regional wall-motion abnormalities, or isolated global hypokinesia without scar, in the absence of dilation. The category fills the space between an apparently normal heart and overt DCM. Its recognition depends largely on magnetic resonance, which reveals scar not detected by echocardiography. NDLVC remains a phenotype and is not equivalent to left arrhythmogenic cardiomyopathy of genetic cause.

Scar should be defined as non-ischemic after coronary artery disease, myocarditis, sarcoidosis and other causes have been considered. Mid-wall septal, subepicardial or ring-like patterns suggest different processes, but none is absolutely specific. Absence of dilation does not imply low risk, particularly when LGE, arrhythmias or genotypes such as DSP and FLNC indicate an electrical substrate. Classification makes it possible to describe this phase without waiting for the ventricle to dilate before legitimizing the diagnosis.

Arrhythmogenic right ventricular cardiomyopathy remains a distinct phenotype characterized by right ventricular structural and functional abnormalities associated with tissue, electrocardiographic, arrhythmic and familial findings. Isolated dilation or dyskinesia is insufficient because volume overload, pulmonary hypertension, congenital heart disease and athletic adaptation can produce an abnormal right ventricle. The 2010 Task Force Criteria improved specificity for the classic form but are less sensitive for left-sided disease. The Padua criteria subsequently broadened description of the biventricular and left-sided spectrum.

The broader term arrhythmogenic cardiomyopathy is used clinically for right-sided, left-sided and biventricular forms in which scar and arrhythmias predominate. However, the ESC classification retains ARVC as one of the five phenotypes and describes left ventricular involvement through NDLVC, DCM or overlapping traits, completed by etiology. This choice separates what is directly observable from a broader pathogenetic category. In clinical communication it is useful to specify both levels, avoiding application of the word arrhythmogenic to every cardiomyopathy capable of causing arrhythmias.

Restrictive cardiomyopathy is defined by restrictive physiology, normal or reduced ventricular diastolic and systolic volumes, and normal wall thickness. Rapid early filling followed by abrupt cessation, a high E wave and elevated pressures reflect poor distensibility, but Doppler parameters depend on rhythm, preload and stage. Ejection fraction may remain preserved because it measures a proportion ejected from a small cavity rather than absolute output. Diagnosis requires exclusion of constrictive pericarditis, in which the impediment to filling is pericardial and may be surgically treatable.

Infiltrative diseases with thickened walls may produce restrictive hemodynamics without morphologically belonging to RCM according to the ESC definition. An amyloid heart with increased wall thickness is described as a hypertrophic phenotype associated with amyloid etiology, despite high filling pressures and reduced distensibility. This apparent terminological rigidity prevents confusion between physiological behavior and a morphological category. In practice, both levels should be reported because physiology drives symptoms and etiology determines treatment.

Endomyocardial forms pose an additional problem. Endomyocardial fibrosis and hypereosinophilic syndrome may obliterate the apices, involve chordae and leaflets, and cause mitral or tricuspid regurgitation, producing a restrictive picture with distinctive anatomical features. Classification should describe endocardial involvement and the cause rather than reducing the case to RCM alone. This precision becomes essential when anticoagulation, immunosuppression or surgery depends on the phase of the endomyocardial process.

NDLVC should not become a new residual category for every incidental LGE finding. Small scars may result from remote myocarditis, procedures, emboli or sarcoidosis and require plausible belonging to a cardiomyopathic process. Amount, distribution, wall motion, arrhythmias and family history determine significance. The category is particularly useful when scar is associated with a genotype or progression, whereas an isolated finding should retain cautious wording.

ARVC is diagnosed through combinations of criteria and not simply by finding a desmosomal variant. A PKP2 carrier without abnormalities remains at risk and is followed, but does not automatically have the phenotype; conversely, a clinically defined form may remain genetically unresolved. The Task Force criteria favor specificity and the classic right-sided form, whereas the Padua criteria include left-sided involvement more explicitly. Reporting which framework was applied makes the assessment reconstructible.

The distinction between RCM and constrictive pericarditis shows why morphological classification cannot disregard physiology. Both produce elevated filling pressures, but accentuated respiratory ventricular interdependence characterizes constriction, in which filling is limited by a noncompliant envelope and dissociation occurs between intrathoracic and intracardiac pressures; in restriction the defect is myocardial. Doppler, septal motion, hepatic veins, pericardial imaging and invasive hemodynamics are integrated. An incorrect category may deny a patient potentially curative pericardiectomy or expose the patient to unnecessary surgery.

Morphological traits, overlap and borderline phenotypes

Left ventricular hypertrabeculation is considered a morphological trait because it may occur in healthy hearts and in physiological or reversible conditions. Criteria based on the ratio between noncompacted and compacted layers or on trabeculated mass identify a morphology, but their specificity decreases when applied outside the population in which they were developed. Athletes, pregnant women and people with high preload may meet them without developing events. Pathological diagnosis requires integration with function, LGE, arrhythmias, thrombi, neuromuscular disease and family history.

The definition of noncompaction cardiomyopathy remains in use when hypertrabeculation and myocardial disease form a clinically meaningful whole, but it should not assume a universal embryologic arrest. Variants in sarcomeric and cytoskeletal genes may be associated with the trait without it always being the main mechanism of dysfunction. A patient with DCM and prominent trabeculation may be more usefully described as DCM with hypertrabeculation than as a separate entity. This wording preserves the dominant phenotype and reduces overdiagnosis.

Isolated right ventricular dysfunction is also a trait that requires a broad differential diagnosis. Pulmonary embolism, pulmonary hypertension, shunts, congenital abnormalities, tricuspid regurgitation, right ventricular infarction and athletic adaptation can produce dilation or hypokinesia. Only after a proportionate assessment should a primary myocardial process be considered, integrating ECG, arrhythmias, CMR and family history. The ARVC label should not be assigned by exclusion on the basis of a right ventricle that is difficult to measure.

Borderline phenotypes require attention to the biological continuum and measurement error. A wall thickness of 14 rather than 15 mm, a volume just above the limit or a minimal area of LGE do not instantly turn normality into disease. Thresholds and z-scores make diagnosis reproducible, but meaning depends on technical quality, distribution, pre-test probability and associated findings. Repeating the examination with better technique or following evolution may be more informative than immediately forcing a category.

Overlaps may result from progression or gene biology. DSP may produce non-dilated left ventricular scar, inflammatory episodes and subsequent biventricular dysfunction; MYH7 may be associated with hypertrophy, dilation or hypertrabeculation; DES may involve conduction, skeletal muscle and restrictive physiology. In these cases, choosing a single label hides information, whereas a hierarchical description identifies the dominant phenotype, secondary traits and etiology. The best classification is the one that preserves clinically relevant data with the fewest assumptions.

Common conditions may overlap as well. Hypertension and obesity increase mass and volumes, coronary artery disease adds ischemic scar, and valvular disease changes loading conditions and chambers. The presence of a genetic variant does not make these factors irrelevant, just as their presence does not erase an inherited substrate. Diagnosis should attribute a plausible share of the phenotype to each component and define which are modifiable. A multifactorial model is often more faithful than the search for a single cause.

Pregnancy and training demonstrate that morphology can adapt to load without intrinsic disease. Increased trabeculation or volume may regress when the stimulus changes, whereas scar, complex arrhythmia or progression independent of load indicates pathology. Reversibility should not be used as the only criterion, because cardiomyopathy itself can show favorable remodeling. Time, context and the overall set of findings distinguish adaptation from disease better than an isolated measurement.

Interobserver variability is particularly important in borderline traits. Apical segmentation, inclusion of trabeculae, the method used for right ventricular volumes and strain software can change the apparent category. A laboratory should use consistent protocols and references, while the clinician should know the uncertainty of the measurement. Discussing a value with false precision, for example as though a one-millimeter difference were biologically certain, changes diagnosis and anxiety without increasing accuracy.

Etiologic, familial and genetic integration

Once the phenotype has been defined, the etiologic pathway uses clues from age, mode of onset, extracardiac features, ECG, tissue and pedigree. Hypertrophy with pre-excitation suggests a different direction from hypertrophy with low voltages; DCM with early block requires a different priority from dysfunction that developed after persistent tachycardia. Pre-test probability determines which tests can produce useful information. An indiscriminate panel does not replace construction of a hypothesis and increases uninterpretable variants.

Genetic testing can transform a descriptive classification into an etiologic diagnosis, but only when the variant has sufficient evidence. Pathogenic and likely pathogenic categories support diagnosis and cascade testing in the appropriate context; a VUS remains a finding to be reassessed and should not be used to separate healthy from at-risk relatives. Segregation may contribute to reclassification, but absence of phenotype in a young carrier does not prove benignity. Incomplete penetrance also requires caution when interpreting pedigrees.

Familial classification is not limited to "familial" or "sporadic." A de novo variant can cause genetic disease without affected relatives, a small family may not show segregation, and early extracardiac death may prevent penetrance from being observed. Conversely, hypertension or shared exposures may create non-monogenic clustering. The pedigree is a dynamic representation that should be updated with age, tests and new diagnoses, not a form completed only once.

Non-genetic causes are classified by mechanism and strength of evidence. Persistent tachycardia followed by recovery after ablation supports arrhythmia-induced cardiomyopathy, whereas reported alcohol consumption without temporality or response to abstinence does not automatically prove alcoholic cardiomyopathy. Drugs, infections, autoimmunity and pregnancy require the same causal rigor. The term "multifactorial" is appropriate when several plausible components interact, but should not become a way to stop the investigation.

Extracardiac involvement broadens classification and may reveal the diagnosis. Neuropathy, nephropathy, muscle weakness, deafness, ocular, hepatic or skin abnormalities link the cardiac phenotype to systemic diseases. Their absence does not exclude forms that begin in the heart, whereas their presence should be confirmed with specific tests. A purely cardiologic classification is insufficient when therapy and prognosis depend on multiorgan burden.

Etiologic diagnosis modifies surveillance even with identical morphology. A patient with an LMNA variant and modest dysfunction may require greater attention to conduction and arrhythmic risk than a recovered tachycardiomyopathy; HCM due to Fabry disease requires renal and neurologic assessment and offers specific therapy; ATTR and AL have completely different hematologic and systemic pathways. This demonstrates why phenotypic classification is necessary but insufficient. Stopping at the first level means giving up the most actionable part of the diagnosis.

A variant should be linked to the phenotype through gene validity, not merely through a computational prediction. Genes proposed in small families may lose support when compared with large populations; truly pathogenic variants may instead be underestimated if the specific mechanism is not considered. Clinician and laboratory should share sufficient phenotypic data without retrospectively adapting every finding to the variant discovered. Genetic classification remains independent of the desire to close the case.

Molecular reclassification may change the position of relatives without immediately changing the proband's phenotype. If a VUS becomes pathogenic, cascade testing becomes possible; if a variant is downgraded, relatives should not continue to be defined as affected on that basis alone. Clinical decisions also supported by imaging or arrhythmias remain valid independently of the variant. Separating phenotypic and genetic axes allows one to be updated without improperly erasing the other.

MOGE(S), clinical staging and limits of classification

The MOGE(S) nomenclature represents five axes: M describes the morphofunctional phenotype, O the organs involved, G the genetic or familial pattern, E the etiology, and S the functional status or stage. The structure makes it possible to code a complex case without compressing every element into a single word. For example, dilated cardiomyopathy with muscle disease, autosomal dominant inheritance and a defined variant can be distinguished from isolated acquired DCM despite sharing the same ventricular image. Granularity facilitates registries and comparison among cohorts.

The limitation of MOGE(S) is its operational complexity. A complete string requires data that may not be available at the first visit and may become difficult to read for those who do not use the system regularly. The solution is not to abandon the multidimensional approach, but to translate it into a structured clinical sentence reporting phenotype, cause, family and phase. The code can accompany this sentence in research settings without replacing it in communication with patients and non-specialist professionals.

Staging adds a temporal dimension. Genotype-positive status without phenotype, subclinical expression, structural disease, symptoms and advanced heart failure are not equivalent categories, even when they share the same etiology. Heart failure stages, NYHA functional class and cardiopulmonary capacity describe complementary aspects but do not summarize arrhythmic or extracardiac risk. A complete classification should therefore prevent a single severity indicator from obscuring the other domains.

Every system reflects the technologies and knowledge of its time. New biomarkers, quantitative imaging and polygenic genetics will probably make boundaries now defined by thresholds more continuous. Overly rigid categories risk hindering recognition of emerging forms, whereas excessively free language reduces reproducibility and scientific comparison. The practical solution is to maintain stable operational definitions while documenting within them findings that do not fit perfectly.

Classification causes harm when it is used to infer therapy or prognosis automatically. Not all hypertrophic phenotypes respond to myosin inhibitors, not all non-dilated scars carry identical risk, and not every trabeculated ventricle requires anticoagulation. Decisions derive from studies conducted in defined populations and should be transferred only when cause, phase and patient characteristics are comparable. The category name opens the decision pathway, but does not constitute its conclusion.

The most durable principle is to distinguish description from explanation. The former should be precise, reproducible and updated; the latter should be supported by proportionate evidence and state what remains unknown. By linking phenotype, etiology, family and time, classification of cardiomyopathies becomes a method of reasoning rather than an inventory of terms. It is this function, more than the permanence of a specific taxonomy, that makes it clinically useful.

Coding can be granular without making the text unreadable if it is applied in layers. The main report presents the dominant phenotype and cause, a subsequent section reports traits, organs and phase, and the MOGE(S) string preserves standardized detail. This hierarchy allows an emergency physician to grasp the problem immediately and a specialist center to recover the full complexity. A code without clinical translation and a sentence without verifiable criteria both fail the communicative function.

Registries and studies should declare which version of the classification was used, because recoding old cohorts with new categories changes denominators and events. Patients once included in DCM may now fall within NDLVC, and isolated noncompaction forms may be treated as traits. Unharmonized temporal comparisons attribute to biology differences produced by nosology. Transparency of criteria is therefore part of the scientific validity of results.