Pathophysiology of dilated cardiomyopathy: from mechanisms to precision medicine

Abstract

Dilated cardiomyopathy (DCM) is a complex disease with multiple causes and various pathogenic mechanisms. Despite improvements in the prognosis of patients with DCM in the past decade, this condition remains a leading cause of heart failure and premature death. Conventional treatment for DCM is based on the foundational therapies for heart failure with reduced ejection fraction. However, increasingly, attention is being directed towards individualized treatments and precision medicine. The ability to confirm genetic causality is gradually being complemented by an increased understanding of genotype–phenotype correlations. Non-genetic factors also influence the onset of DCM, and growing evidence links genetic background with concomitant non-genetic triggers or precipitating factors, increasing the extreme complexity of the pathophysiology of DCM. This Review covers the spectrum of pathophysiological mechanisms in DCM, from monogenic causes to the coexistence of genetic abnormalities and triggering environmental factors (the ‘two-hit’ hypothesis). The roles of common genetic variants in the general population and of gene modifiers in disease onset and progression are also discussed. Finally, areas for future research are highlighted, particularly novel therapies, such as small molecules, RNA and gene therapy, and measures for the prevention of arrhythmic death.

Introduction

Dilated cardiomyopathy (DCM) is one of the four classic cardiomyopathy phenotypes and is defined as dilatation and systolic dysfunction of the left ventricle, with or without biventricular involvement, in the absence of coronary artery disease or other secondary causes, such as systemic hypertension or valvular heart disease¹. DCM is the most frequent non-ischaemic cause of heart failure (HF), a principal indication for heart transplantation in young patients, and a major cause of sudden cardiac death (SCD). The prevalence of DCM has been estimated at approximately 1 in 250 adults².

Advances in knowledge over the past decade, in particular the recognition that different pathogenic variants in the same cardiomyopathy gene can cause different phenotypes, has blurred the distinction between non-hypertrophic phenotypes and expanded the area of overlap between DCM and arrhythmogenic cardiomyopathy (ACM). Moreover, as the understanding of aetiological mechanisms has progressed, strict separation into genetic and non-genetic aetiologies is now thought to be overly restrictive because multiple interactions occur between the genetic background and exposure to environmental triggers (Fig. 1). Future models that merge multidimensional genetic, environmental and behavioural information will lead to more comprehensive assessment of patients with DCM³, including more precise prognoses. Moreover, contemporary research is focused on novel strategies of treatment that deviate from classical antagonism of the neurohormonal axis and aim to deliver precisely targeted therapies, including gene therapy and modulators of primary and secondary disease pathways⁴. The first guidelines for the management of cardiomyopathies were published in Europe in 2023 (ref. 5), attesting to the remarkable advances in the field achieved by the cardiology research community. These European guidelines specify cut-off points for left ventricular (LV) dilatation (LV end-diastolic diameter >58 mm in men and >52 mm in women, and LV end-diastolic volume index ≥75 ml/m² in men and ≥62 ml/m² in women) and LV systolic dysfunction (LV ejection fraction (LVEF) <50%), but also included a new pathological entity characterized by a non-dilated LV cardiomyopathy phenotype, which includes LV systolic dysfunction in the absence of LV dilatation⁵.

Fig. 1 ∣. Complex interactions in the pathophysiology of dilated cardiomyopathy.

Multiple interactions occur between environmental (red boxes) and genetic (yellow boxes) factors, with monogenic or polygenic architectures. Titin-truncating variants (TTNtv) are a model of the ‘two-hit’ hypothesis, whereby genetic predisposition to dilated cardiomyopathy can mediate different responses depending on environmental factors and, reciprocally, that environmental exposure might influence gene expression. Solid lines indicate strong evidence, whereas dashed lines indicate areas still to be explored.

In this Review, we provide a comprehensive overview of DCM pathogenesis, including genetic and acquired causes and the complex interplay between genetic abnormalities and triggering environmental factors (the ‘two-hit’ hypothesis). We also explore approaches to individualized, precision medicine and discuss current and future therapeutic strategies and avenues for research.

Genetic causes of DCM

Genetic epidemiology

Genetic cardiomyopathies have classically been considered to be monogenic, Mendelian disorders, most frequently with autosomal dominant transmission and incomplete penetrance. The genetic architecture of DCM is the most heterogeneous among the primary cardiomyopathy phenotypes⁶. Variants implicated in DCM encode components of every cardiomyocyte compartment and with various functions². Before the advent of novel sequencing techniques, the definition of causality was limited to a paucity of genes that were tested singularly in pioneering studies². In these early studies, the proportion of DCM with causative genetic variants was fairly low (20–35%), particularly considering the strong selection bias of the populations studied⁷. The advent and spread of next-generation sequencing has increased accessibility to genetic testing and has had a substantial effect on the diagnosis and management of DCM, as indicated in the European guidelines⁵.

In 2012, a landmark study found that titin-truncating variants (TTNtv) are the most frequently identified cause of genetic DCM (approximately 25% of familial cases and 18% of sporadic cases), highlighting the central role of titin in normal sarcomere function and marking a turning point for clinicians and researchers⁸. After this discovery, the diagnostic yield for genetic testing expanded⁹, although some bias in the referral populations studied was still present. In a contemporary cohort of patients with DCM from two referral centres for the management of genetic cardiomyopathies, the proportion of various causative variants increased up to 40%¹⁰. Although TTNtv are the main genetic cause of DCM (10–20% of patients), variants in sarcomeric genes (specifically MYH7 and TNNT2) are responsible for up to 10% of genetic forms of DCM. The gene encoding lamin A/C (LMNA) is also a major contributor to the genetic epidemiology of DCM, accounting for the disease in up to 6% of patients, whereas other gene variants are less frequent (<3%)^10-13. Table 1 shows the phenotypes associated with genes that have a definitive or strong causal link with DCM.

Table 1 ∣.

Phenotypes for genes with a definitive or strong causal association with DCM^18,21

Gene	Protein	Classification	Age at onset (years)	Women (%)	Prevalence in DCM (%)10,13,46	Phenotypes	Arrhythmic risk10,46	Heart failure risk10,46	LVRR44-46	Additional notes
BAG3	BCL2-associated athanogene 3	Definitive	20–49	38	1–2	DCM	+	+++	+	Earlier onset of DCM and higher HF risk in men than in women, low rate of appropriate ICD shocks181
DES	Desmin	Definitive	35–60	35	1–2	DCM	−	+++	−	Limited phenotype and outcome information182
DSP	Desmoplakin	Strong	20–49	69	1–3	DCM, ACM	+++	++	++	Biventricular involvement, hot phases, ‘ring-like’ LGE distribution, ACM183
FLNC	Filamin C	Definitive	25–59	47	2	DCM, ACM	+++	+	++	Higher risk in men than in women, ‘ring-like’ LGE distribution162
LMNA	Lamin A/C	Definitive	25–55	43–56	2–6	DCM, ACM	++++	++++	+	Higher risk in men than in women, frequent AV conduction abnormalities and supraventricular arrhythmias, rare RCM 28,126,184-187
MYH7	Cardiac myosin heavy chain 7	Definitive	20–49	42	1–5	DCM	+	+++	+	16% onset in patients aged <18 years, frequent HCM, LVNC, rare ACM186,187
PLN	Phospholamban	Definitive	25–55	57	<1	DCM, ACM	+++	++	−	Most evidence for p.Arg14del founder variant158
RBM20	RNA-binding motif protein 20	Definitive	20–49	NA	1–2	DCM, ACM	+++	+	−	Highly arrhythmogenic, highly penetrant188-190
SCN5A	Sodium voltage-gated channel, α-subunit 5	Definitive	15–25	27	1–2	DCM, ACM	++	+	−	Frequent supraventricular arrhythmias and AV conduction abnormalities, rare ACM 191,192
TNNC1	Troponin C	Definitive	25–55	28	<1	DCM	+	++	++	Evidence available for sarcomeric gene clusters18
TNNT2	Troponin T	Definitive	25–55	28	1–6	DCM	+	++	++	Evidence available for sarcomeric gene clusters10,46
TTN	Titin	Definitive	40–69	29–41	10–20	DCM, LVNC	+	+++	++++	Higher risk in men than in women, rare ACM and myocarditis 8,27,32,193-196

−, no association or predisposition; +, weak association or predisposition; ++, moderate association or predisposition; +++, strong association or predisposition; ++++, very strong association or predisposition; ACM, arrhythmogenic cardiomyopathy; AV, atrioventricular; DCM, dilated cardiomyopathy; HCM, hypertrophic cardiomyopathy; HF, heart failure; ICD, implantable cardioverter–defibrillator; LGE, late gadolinium enhancement; LVNC, left ventricular non-compaction cardiomyopathy; LVRR, left ventricular reverse remodelling; NA, not available; RCM, restrictive cardiomyopathy.

Defining and interpreting causal gene variants

The complexity of interpreting genetic tests for DCM is considerable for various reasons, including the growing number of positive tests, variable gene penetrance and disease expressivity, the heterogeneous landscape of candidate genes with partial or inconclusive causality, and the modest proportion of carriers in the general population (for example, 2.64% of participants in the UK Biobank with TTNtv¹⁴), which carries potential prognostic implications. Establishing the pathogenicity of identified variants is challenging. In 2015, the American College of Medical Genetics and Genomics (ACMG) updated its standards and guidelines for the interpretation of genetic testing, which define criteria and terminology¹⁵ (Box 1). The ACMG criteria are based on a combination of genetic, bioinformatic and medical literature data, and are used to assigned gene variants to one of five categories (benign, likely benign, uncertain significance, likely pathogenic and pathogenic). In addition, a causative role in DCM is considered for many pathogenic and likely pathogenic variants in the ClinGen database. Nevertheless, interlaboratory variability has been reported and can lead to misclassifications^16,17. Moreover, as variant reporting increases, the evidence on causality of variants can change and, for this reason, pathogenicity according to ACMG criteria should be periodically reassessed¹⁸. Notwithstanding variant reclassifications, clinicians are responsible for interpreting the results in the context of the clinical phenotype, which in many cases can extend beyond their expertise and should ideally be done by those with sufficient training at specialized centres¹⁹.

Box 1 ∣. ACMG criteria for variant classification.

Definitions and important aspects of variant classification from the American College of Medical Genetics and Genomics (ACMG)¹⁵.

Variant classification

• Pathogenic

• Likely pathogenic

• Uncertain significance

• Likely benign

• Benign

Criteria

• Supporting

• Moderate

• Strong

• Very strong

Data domains

Of note, not all domains contribute data in each case.

• Population data

• Computational (in silico) prediction

• Functional data

• Segregation data

• De novo status

• Allelic data

• Other (subjective, based on specific cases)

Genetic testing in patients with dilated cardiomyopathy

• Pathogenic and likely pathogenic variants are clinically actionable, meaning that they can be used for predictive testing in relatives, for clinical decision-making or both.

• Variants of uncertain significance are defined when criteria for classifying pathogenic or benign variants are not met or are contradictory.

• The level of evidence for variants can change over time; therefore, periodic reassessment of variant significance is strongly advised, particularly for variants of uncertain significance.

• Selection of appropriate gene panels must be balanced between a reasonable probability of causal association between gene variants and disease (restrictive panels) and a low probability of missed diagnosis (extensive panels).

• Full phenotyping and clinical history of patients and relatives are vital for the interpretation of gene variants.

The high number of candidate genes involved in the pathogenesis of DCM has led to the inclusion of 80–120 genes in extended gene panels. However, many of these genes lack a strong association with the disease, resulting in the identification of variants of uncertain significance, making the interpretation of genetic test results in clinical practice even more challenging²⁰. In two studies, one with a semiquantitative approach¹⁸ and the other with a case–control design²¹, points were assigned according to clinical gene evidence and experimental evidence. In the datasets from each study, 19 of 51 genes¹⁸ and 12 of 56 genes²¹, respectively, were identified as having a robust association with DCM (Table 1), which could reasonably represent the basis for diagnostic genetic panels. Conversely, the complexity of overlapping phenotypes in cardiomyopathies can lead to missed diagnosis if an overly restrictive panel is selected. In a large population of patients referred for either inherited cardiomyopathies or arrhythmias, comprehensive genetic testing for these conditions increased the yield of diagnosis by 14.4% compared with disease-specific testing based on physician clinical diagnosis⁹. Improved diagnostic yield could outweigh the burden of uncertain results in some circumstances. Furthermore, clear definition of the goals of genetic testing and identification of patients who have an increased probability of positive results (Box 2) could further improve the diagnostic yield. In a multicentre study conducted in Spain and externally validated in a European cohort, the use of a simple multivariable score enabled a positive test to be predicted in 79% of carriers fulfilling all the identified criteria²². The variables used to estimate the probability of a positive result were family history of DCM, the presence of low voltage in peripheral electrocardiographic leads, concomitant skeletal myopathy, and the absence of hypertension and left bundle branch block²². This score, as well as the points listed in Box 2, could guide clinical cardiologists in the identification of candidates for genetic testing, thereby optimizing resource allocation.

Box 2 ∣. Goals of genetic testing in DCM.

Whom to test?

• Probands with dilated cardiomyopathy (DCM), with potential clinical or familial application of results (see below)

• First-degree relatives, for identification of carrier status and activation of genetic cascade screening and early diagnosis (see Fig. 3), for confirmation of variant pathogenicity (cosegregation) or for research purposes

• Balance between availability of resources (accessibility and costs), likelihood of positive testing and applicability of results

Factors indicating an increased likelihood of positive testing or suspicion of malignant genotypes (see Fig. 2)

• Absence of alternative or contributing causes (such as autoimmune diseases, viral infections, arrhythmias, endocrine disorders, exposure to toxins and pregnancy)

• Familial DCM, or undefined left ventricular or biventricular dysfunction, at a young age

• Family history of sudden cardiac death or life-threatening arrhythmias, hypokinetic arrhythmias or unexplained death in childhood, adolescence or early adulthood

• Extracardiac (including myopathy) involvement

• Sudden cardiac arrest or sustained ventricular tachycardia

• Specific electrocardiographic phenotypes — low voltages on peripheral leads, negative T waves, atrioventricular block (any degree) or absence of left bundle branch block

• High arrhythmic burden (for example, multiple ventricular ectopic beats, non-sustained ventricular tachycardia)

• Specific echocardiographic phenotypes — right ventricular involvement, regional akinesia/dyskinesia or aneurysm, or a non-dilated phenotype

• Specific patterns of late gadolinium enhancement on cardiac MRI (such as subepicardial ‘ring-like’ scar)

• Madrid genotype score of ≥2 points (probability of positive test 37% higher than in the general populations with DCM)²²

Factors supporting the clinical or familial application of genetic test results

• Young age of patient

• Low comorbidity burden (low competing risk of morbidity events or mortality due to other causes)

• To determine the need for prophylactic cardioverter–defibrillator implantation

• Presence of children or siblings (cascade genetic screening)

• Planned pregnancy

When to test?

• At DCM diagnosis if fulfilling the above criteria

• According to patient’s and family’s decisions

• With changes in disease status (cardiac arrest or life-threatening arrhythmias, hypokinetic arrhythmias or need for a pacemaker, or recurrent syncope)

• With new emergence of gene-specific signs of disease on electrocardiography or imaging

• With phenotype manifestation in relatives

Gene-elusive disease is frequent in selected populations with DCM, despite the high prevalence of familial cases, suggesting that alternative mechanisms are involved. The extensive information derived from genome-wide association studies, gene-to-gene interactions and the emerging field of polygenic risk scores could contribute to the identification of polygenic traits in patients with gene-elusive DCM that does not fit the classic monogenic model of transmission^23-25.

Sex-related differences

Sex-related differences in genetic DCM are emerging. Women are less represented than men in all the major genetic DCM cohorts, suggesting a lower penetrance of the disease^10,12. In carriers of BAG3 variants, male sex is associated with earlier onset of disease and worse prognosis than female sex and by a higher incidence of HF-related (but not of arrhythmic) events²⁶. Worse outcomes have been reported for male carriers of TTN and LMNA variants than for female carriers and, in two separate studies, male sex was strongly associated with an increased risk of life-threatening arrhythmias^27,28. Indeed, male sex is one of the risk factors for ventricular arrythmias and SCD in carriers of LMNA variants, and is considered in the indications for an implantable cardioverter–defibrillator (ICD)²⁹. In filamin C (FLNC) variant cardiomyopathy, weak evidence exists for a numerically lower risk of major cardiovascular events in women than in men³⁰, but larger studies are strongly advocated to define sex-related differences in arrhythmic burden and the risk of SCD. In a single-centre study in the UK published in 2024, women presented with characteristics of less severe disease, including higher LVEF, less LV remodelling and a lower cardiac fibrotic burden than men, but had higher rates of HF events and cardiovascular death in the first 2 years of follow-up³¹. Frequency and distribution of genetic variants were similar between the sexes. Such findings highlight a need for focused studies designed to define the mechanisms underlying the sex-related differences in the risk of HF and arrhythmic events and to resolve residual uncertainties.

Towards genotype-centred management

Increasing awareness that the distinction between traditional phenotypes is insufficient to categorize the multiplicity of cardiomyopathy presentations, particularly DCM, could herald a new approach in which the genotype represents the core of clinical management. Indeed, the frequent phenotypic overlap between DCM and other cardiomyopathies, particularly ACM, which is often caused by variants in DSP or FLNC, can preclude the precise categorization of individual patients. Evidence also suggests that lifetime disease penetrance can be influenced by the variant carried. For example, TTN-related DCM is more likely to present later in life than DCM caused by other genotypes³². Various DCM genotypes can also present with different phenotypes (Fig. 2) and risk of disease progression and arrhythmic and non-arrhythmic events (Table 1). Although LMNA was first identified as a high-risk gene for early HF-related outcomes or SCD in patients with DCM, additional genes such as DSP, FLNC, PLN and RBM20 have subsequently been associated with a worrisome burden of life-threatening arrhythmias in these patients (Fig. 3). For the most frequently observed genotypes, risk scores have been proposed and are under investigation. However, with the exception of LMNA, studies of these genotypes are generally biased by limited population size and the lack of validation cohorts²⁸.

Fig. 2 ∣. Genotype–phenotype correlations in dilated cardiomyopathy.

Late gadolinium enhancement (LGE) on cardiac MRI and electrocardiographic findings in four different genotypes of dilated cardiomyopathy. a, A carrier of a pathogenic variant in FLNC, with a typical subepicardial ring-like pattern of LGE on MRI, as well as low voltages on peripheral leads and inferolateral negative T waves on the electrocardiogram. b, A carrier of a pathogenic variant in TTN, with no specific LGE distribution and minor, non-specific repolarization electrocardiographic abnormalities. c, A carrier of a pathogenic variant in DSP, with a typical subepicardial ring-like pattern of LGE on MRI, as well as low voltages on peripheral leads, ventricular ectopic beat, QRS fragmentation and flat lateral T waves on the electrocardiogram. d, A carrier of a pathogenic variant in LMNA, with transmural septal LGE on MRI as well as first-degree atrioventricular block, intraventricular conduction delay and multiple ventricular ectopic beats on the electrocardiogram.

Fig. 3 ∣. Life-threatening ventricular arrhythmias and HF-related events in DCM.

Graphs showing survival free from sudden cardiac death (SCD) or major ventricular arrhythmias (MVA) (panel a) and heart failure-related (HF) death or heart transplantation (HT) (panel b) for the four genotypes of dilated cardiomyopathy (DCM) in the Trieste, Italy, and Denver, USA, genetic cardiomyopathy registries¹⁰: DSP, FLNC, LMNA and TTN. The risk of SCD or MVA is similar in DSP, FLNC and LMNA variant groups and lower in the TTN variant group. Carriers of LMNA variants have the highest risk of HF-related death or HF. Data extracted from ref. 10.

The role of genetic testing in DCM has traditionally been limited to guiding screening programmes in families of genotype-positive probands. If a pathogenic or likely pathogenic variant is found, family cascade screening should be activated^33,34, with the goal of detecting early, overt disease in first-degree relatives, whereas non-carriers can be discharged from periodic screening³⁵ (Fig. 4). Genotype-positive, phenotype-negative relatives of patients with DCM should undergo periodic, systematic assessment, including electrocardiography and echocardiography. In addition, advancing knowledge of imaging strategies such as speckle-tracking echocardiography and cardiac MRI for the identification of preclinical disease could help to identify candidates for early treatment, prevent overt disease manifestation and reduce the lifetime risk of arrhythmias and HF^35-37. However, early treatment in carriers of genetic variants associated with DCM, without disease manifestations, is not currently supported by the available evidence. The EARLY-GENE trial³⁸ is an ongoing randomized study of candesartan in genotype-positive, phenotype-negative individuals to prevent progression to DCM. The treatment of asymptomatic LV systolic dysfunction with angiotensin-converting enzyme inhibitors or β-blockers has previously been shown to improve outcomes (mortality or incidence of HF)^39,40, but idiopathic DCM was under-represented in these studies and they were not specifically designed for genetic DCM.

Fig. 4 ∣. Cascade screening in DCM.

Flow chart proposal for genetic testing and subsequent management of probands with dilated cardiomyopathy (DCM) and their first-degree relatives. Genetic testing and clinical assessment also apply to first-degree relatives of each affected individual or carrier of a pathogenic or likely pathogenic variant within the family.

Advancing knowledge and technology can help clinicians to identify the preliminary signs of variant penetrance in relatives at risk of DCM. Reduced global longitudinal strain is the strongest parameter associated with the likelihood of disease expression and adverse prognosis. Deformation imaging, assessed by speckle tracking echocardiography, is more sensitive than conventional echocardiography for the quantification of systolic function and, therefore, allows the early identification of subtle abnormalities in global longitudinal strain⁴¹. Importantly, according to the guidelines⁵, isolated LV dilatation with normal LVEF is sufficient for a diagnosis of DCM in a relative of a proband with DCM in the presence of other mild non-diagnostic abnormalities or if a familial causative variant is identified⁵. In addition, the guidelines include a class IB recommendation for genetic testing for phenotype-positive patients, primarily for prognostic and therapeutic stratification⁵.

A treatment response, as expressed by the improvement or recovery of LV function (LV reverse remodelling (LVRR)), with guideline-directed medical therapy occurs in up to 40% of patients with DCM within 2 years after diagnosis⁴². However, carriers of DCM-associated variants seem to respond less effectively to anti-neurohormonal medications than non-carriers^43,44. Variability in the rate of LVEF recovery has also been demonstrated across different genotypes. Carriers of TTNtv consistently have the highest rates of LVRR, exceeding 50% in all the main studies^12,44,45. In carriers of myosin motor sarcomeric gene variants, LVRR was between 23% and 28%; the likelihood of LVRR is even lower in carriers of LMNA pathogenic variants^12,44,45. The reported independent, inverse association between cytoskeleton Z-disc gene variants and LVRR, which was not confirmed in other cohorts, was probably related to the high prevalence of malignant FLNC variants in this subgroup⁴⁵. In patients with complete LVEF recovery, guideline-directed medical therapy must be continued indefinitely, given that a relapse in LV dysfunction and the development of HF are highly likely after treatment withdrawal⁴⁶.

The contribution of positive genotype to the prognosis in DCM has long been controversial⁴⁷. In a study in 487 patients with DCM, all-cause mortality was similar between those with an identified gene variant and those without, but a trend towards a higher incidence of HF-related outcomes and life-threatening arrhythmias was observed in those with a gene variant¹⁰. A competing risk survival analysis started from birth showed an even greater association with the arrhythmic outcome. Furthermore, a multicentre analysis of 1,005 probands with DCM from 20 centres in Spain supported the association between variant carrier status and outcome, with an increased incidence of major adverse cardiac events (MACE), end-stage HF and malignant ventricular arrhythmias in carriers than in non-carriers¹². An association between advanced disease and more frequent identification of pathogenic or likely pathogenic variants in DCM also emerged when patients were stratified according to the severity of HF, as defined by the need for long-term mechanical circulatory support or heart transplantation⁴⁸. Shifting the focus from phenotype to genotype can also help to predict patient outcomes. In 281 carriers of pathogenic or likely pathogenic variants with heterogeneous non-hypertrophic phenotypes (80% with DCM), genotype-based classification, but not phenotype-based classification, was predictive of life-threatening arrhythmias⁴⁹.

Acquired causes of DCM

In addition to pathogenic or likely pathogenic gene variants, DCM can also be caused by several acquired or ‘secondary’ aetologies⁵⁰. However, the pathophysiological overlap between modifiable or environmental insults and the genetic substrate is complex and remains largely unclear, but is thought to involve inflammation and neurohormonal mechanisms. Non-genetic causes can interact with genetically determined structural abnormalities, as demonstrated in the so-called two-hit hypothesis by the high prevalence of TTNtv in patients with DCM attributed to chemotherapy, excessive alcohol intake or pregnancy. In this section, we focus on some of the important non-genetic causes of DCM, including autoimmunity, viral infections, exposure to toxins, endocrine disorders, pregnancy and arrhythmias.

Autoimmunity

The role of autoimmunity in the pathogenesis of DCM has been extensively investigated. Autoantibodies have been detected in the serum of up to 60% of patients with DCM⁵¹, directed against various heartspecific and muscle-specific self-antigens, including cardiac myosin heavy chain isoforms, β-adrenergic receptors, M₂ muscarinic acetylcholine receptors and cardiac troponins^52-54. Some autoantibodies have been proposed to have a pathogenic role in DCM^55,56 or to be related to a poor prognosis, arrhythmias and SCD^57-59. Furthermore, first-degree relatives of patients with DCM who had asymptomatic LV dilatation showed a higher frequency of circulating anti-heart autoantibodies and a higher 5-year rate of progression to DCM than control patients with non-inflammatory heart disease⁶⁰. However, despite promising evidence in animal models^61,62, the role of autoantibodies in the context of the highly heterogeneous DCM phenotype in humans remains controversial. This lack of evidence could be partly due to the largely unknown concentration, acuity and time course of autoantibodies in the myocardium. Moreover, the complex interplay between genetics, and environmental and epigenetic modifiers influences the autoimmune response that promotes LV remodelling and leads to DCM. Further studies are necessary to determine whether autoantibodies should be considered diagnostic markers of DCM rather than causative agents.

Viral infections

The role of bacteria (such as Borrelia burgdorferi or Chlamydia pneumoniae), protozoa and fungi in the pathogenesis of DCM is not well understood. Conversely, viral infections are known to cause myocarditis in humans and in animal models^63,64, with a variable and not easily estimated rate of progression to DCM. The most frequently cardiotropic viruses reported in endomyocardial biopsy samples in Europe and the USA are erythroparvoviruses, followed by enteroviruses (including coxsackievirus groups A and B), adenoviruses, influenza viruses and human herpesviruses⁵¹. However, as parvovirus B19 (B19V; a type of erythroparvovirus) has also been reported in non-inflammatory cardiac disease and in healthy hearts⁶⁵, the presence of its genome could indicate a latent virus without replicative activity or a pathogenic role⁶⁶. Moreover, >500 viral copies per microgram of DNA is the current clinically relevant threshold for B19V; above this value, B19V might maintain myocardial inflammation which could possibly evolve to a DCM phenotype⁶⁷. Assessing the transcriptional activity of B19V, viral load, localization of the infection and genetic background of the patient can help to determine the clinical relevance of the infection and whether antiviral therapies might be beneficial⁶⁸.

Viral particles actively replicate in the myocardium and can either induce direct myocardial and endothelial injury or trigger an uncontrolled immune response by molecular mimicry. Ineffective viral clearance, exacerbated by chronic inflammation, has been implicated in the progression from myocarditis to ventricular dilatation and DCM⁶⁹, and has been related to deteriorating cardiac function⁷⁰. However, viral infections were not associated with decreased survival in multivariate analyses⁷¹.

At present, in myocarditis, which might evolve to DCM, the presence of viral genome is considered to be a pathological finding, except in cases in which <500 viral copies per microgram of B19V DNA load are detected⁷². Endomyocardial biopsy is the diagnostic gold standard for detecting viral genome, and expert recommendations emphasize the need to rule out active infection before considering immunosuppressive therapy^51,73-76. However, the use of endomyocardial biopsies is limited due to the requirement for specialized medical teams capable of performing the procedure safely. Consequently, endomyocardial biopsies are recommended primarily in life-threatening situations (such as for treatment-refractory HF or malignant ventricular arrhythmias) or in rare and controversial clinical settings (such as for ‘hot-phase’ cardiomyopathies or chronic troponin release). Although some studies have demonstrated short-term⁷⁷ and long-term⁷⁸ benefits of immunosuppressive therapy on cardiac function in patients with virus-negative myocarditis, its effects on progression to LV dilatation and on survival are still debated^72,79-81. Conversely, some evidence from endomyocardial biopsy samples has been found for the clearance of adenovirus or enterovirus genomes by IFNβ⁶⁹. In selected patients with herpesvirus infection and myocarditis, treatment with aciclovir, ganciclovir or valaciclovir can be considered to prevent progression to DCM^51,82. However, no antiviral therapies have been definitely shown to improve clinical end points in patients with acute myocarditis or DCM.

Exposure to toxins

Alcohol.

Excessive alcohol intake is one of the leading causes of DCM. Alcoholic cardiomyopathy occurs in 1–2% of all individuals who consume alcohol heavily and predominantly in men aged 30–55 years, and is correlated with daily intake and duration of consumption³⁴. Although the threshold of alcohol intake remains unknown, with probably protective or adverse interactions of genetic and environmental factors, 80–90 g of alcohol (approximately 10 units) per day for 5 years is often cited as the level of consumption required to cause DCM^83,84. The pathogenesis of cardiac damage in alcoholic cardiomyopathy includes myocardial ischaemia from increased oxygen consumption, mitochondria damage, contractile impairment and modified calcium homeostasis. Chronic and excessive alcohol consumption was associated with reduced cardiac contractility in mouse models⁸⁵ and with subclinical cardiac remodelling and impaired diastolic function detected by echocardiography in humans^86-88.

As alcoholic cardiomyopathy is clinically and histologically indistinguishable from idiopathic DCM, it is a diagnosis of exclusion, based mainly on the clinical history reported by the patient. Moreover, frequent concomitant atrial fibrillation and arterial hypertension in patients with alcoholic cardiomyopathy make the diagnosis even more challenging⁸⁹. Strict adherence to total abstention from alcohol is recommended in patients with alcoholic cardiomyopathy, as this condition is thought to be reversible⁹⁰, although the clinical data supporting this reversibility are limited.

Cocaine.

Cocaine use has been associated with the development of DCM via ischaemia-independent mechanisms⁹¹. Cocaine seems to induce a persistent hyperadrenergic state and inflammatory response, leading to the release of catecholamines and cardiomyocyte necrosis⁹². Robust data exist for the relationship between cocaine use and acute coronary syndrome⁹³ and LV systolic and diastolic dysfunction⁹⁴. However, a meta-analysis has suggested that, in patients without acute coronary syndrome, cocaine use does not significantly affect LVEF⁹⁵. Additional specific studies are, therefore, needed to clarify the role of cocaine in the pathogenesis of DCM.

Cancer therapies.

Anthracycline-based chemotherapeutic agents (including doxorubicin, epirubicin and idarubicin) used in the treatment of cancer can induce DCM in a dose-dependent and cumulative manner. Cardiac damage can occur in the early phase of chemotherapy or many years after treatment, and therapeutic advances with major improvements in long-term survival have led to an increasing number of patients developing late signs of cardiotoxicity⁹⁶. Moreover, the concomitant use of adjuvant therapies, such as trastuzumab, combined with other risk factors increases the risk of LV dysfunction and HF^97,98.

As for other toxins, anthracycline can provoke multiple effects on cardiomyocytes, including dysregulation of calcium homeostasis, apoptosis and mitochondrial dysfunction^99-101. An integrated approach — comprising clinical assessment, measurement of cardiac biomarkers and serial echocardiography — is recommended to promptly identify symptomatic or asymptomatic LV dysfunction^96,102. The prognosis of chemotherapy-induced cardiomyopathy is also driven by the incidence of non-cardiac events, and identifying patients who can complete chemotherapy protocols despite developing LV dysfunction is crucial to reducing cancer-related mortality¹⁰³. When LV dysfunction is attributed to chemotherapy, the initiation of cardioprotective medications, such as β-blockers and angiotensin-converting enzyme inhibitors, is generally recommended. However, randomized trials of neurohormonal inhibitors for the prevention of anthracycline cardiotoxicity have shown modest or no effects on LVEF and chronic myocardial injury^104,105. Broad consideration of symptoms, cancer prognosis, alternative cancer treatment, possible adverse drug reactions and drug–drug interactions should be discussed with the oncologist and the patient when formulating a treatment plan¹⁰².

Immune checkpoint inhibitors are another class of cancer treatment that warrants attention from cardiologists. These agents use T cell regulatory pathways to inhibit antitumour T cell activation and are used in a growing number of malignancies. Direct cardiomyocyte injury by disinhibited T cells can lead to rare, but increasingly concerning, complications such as fatal myocarditis^106,107. Clinical presentation can vary from asymptomatic ‘smouldering’ to fulminant myocarditis, and endomyocardial biopsy is considered the gold-standard technique for diagnosis and personalized management⁹⁶.

Endocrine disorders

Endocrinopathies — including diabetes mellitus, phaeochromocytoma, and thyroid and growth hormone disorders — are a rare cause of LV dysfunction, suggesting that hormones could also be involved in the pathogenesis of DCM¹⁰⁸. Importantly, men with genetic DCM seem to have a greater risk of systolic dysfunction, atrial and ventricular arrhythmias and MACE than women^{27,28,30,109,110}. Opposite roles for androgens and oestrogen on cardiac vascular endothelial cells, fibroblasts and cardiomyocytes have been demonstrated in vitro and in animal models¹¹¹.

Peripartum cardiomyopathy

Peripartum cardiomyopathy (PPCM) has been linked to various (as yet unproven) mechanisms, including nutritional deficiencies, viral myocarditis and autoimmune processes^112-114, triggered by the hormonal changes that occur in late pregnancy and the postpartum period, and the haemodynamic stress of pregnancy and labour. After excluding pre-existing cardiac conditions, the majority of women with PPCM are diagnosed in the first month after delivery¹¹⁵. This delay in diagnosis is attributed to symptoms, such as oedema and dyspnoea, which are common in late-stage pregnancy. PPCM has been linked to a high rate of adverse outcomes, but affected patients are more likely to recover than patients with other forms of DCM. Recovery often occurs within 3–6 months of diagnosis¹¹⁶.

Arrhythmia-induced cardiomyopathy

Persistent high-rate atrial fibrillation and other arrhythmias, including atrial tachycardia and frequent premature ventricular contractions, have long been recognized as triggers for DCM, with potential reversal of LV dysfunction in treated patients^117-120. Indeed, chronic rapid pacing causes progressive LV remodelling and HF in animal models¹²¹. However, studies in which ablation of atrial fibrillation was compared with medical rate control suggested that high ventricular rate is not the only mechanism involved in arrhythmia-induced cardiomyopathy (AiCM)^122,123. Patients with AiCM generally have a smaller LV end-diastolic diameter, shorter intrinsic QRS duration and often absent or minimal scar burden on cardiac MRI compared with patients with other forms of DCM and concomitant tachyarrhythmias as the clinical consequence of the disease. Additional, focused research is needed to address the clinical challenge of distinguishing between AiCM and DCM with associated tachyarrhythmias early in the disease course^120,124. Currently, a diagnosis of AiCM is made a posteriori by complete or near-complete recovery of LV systolic function within 6 months of achieving rate and rhythm control, and a rapid new decline in LVEF in cases of arrhythmia recurrence¹²⁵. However, even in the context of AiCM, clinicians should not ignore the importance of genetic background. Atrial and ventricular arrhythmias, particularly in individuals aged <60 years, could be early phenotypic features of DCM indicating high-risk genotypes, such as laminopathies¹²⁶.

The role of inflammation

Systemic and myocardial inflammation is a major pathophysiological contributor in genetically determined, genetically susceptible and non-genetic DCM (Fig. 5), with chronic inflammatory cell infiltration revealed on endomyocardial biopsy samples^82,127-129. Mechanisms involved in the pathogenesis of DCM, such as haemodynamic overload, oxidative stress and endothelial dysfunction, can cause or contribute to myocardial injury. The consequent inflammatory response for tissue repair can exacerbate cardiomyocyte damage. Both innate and adaptive immune cells contribute to the progression to myocardial failure^130,131. These cells can be activated by Toll-like receptors expressed on cardiomyocytes, as a reaction to damage-associated molecular patterns (that is, endogenous molecules) or pathogen-associated molecular patterns (that is, exogenous microbial products), stimulating the NACHT, LRR and PYD domains-containing protein 3 (NLRP3) inflammasome and the subsequent so-called IL-1 cascade. Simultaneously, innate immune cells expand adaptative immunity through antigen presentation on major histocompatibility complex class II molecules present on B cells and T cells¹³². Release of cytokines, such as IL-1β, IL-4, IL-17A, IL-33 and transforming growth factor-β1 (refs. 133,134), can cause direct tissue damage, but also create a vicious cycle that perpetuates the immune responses, leading to collagen deposition, fibrotic scar tissue and, ultimately, progression of ventricular dilatation¹³⁵. This process brings into play complex alterations, including increased cardiac workload, hypertrophy and cardiomyocyte apoptosis, as well as changes in the extracellular matrix and expression of fetal genes, proteins and myofibroblasts.

Fig. 5 ∣. Simplified mechanisms of inflammation in dilated cardiomyopathy.

Damage-associated molecular patterns (DAMPs; endogenous molecules) and pathogen-associated molecular patterns (PAMPs; exogenous microbial products), acting via specific receptors and pathways, can trigger and promote inflammatory responses related to the activation of the NACHT, LRR and PYD domains-containing protein 3 (NLRP3) inflammasome in cardiomyocytes. This mechanism triggers an excessive inflammatory response (the IL-1 cascade), mediated by innate and adaptive immune cells, contributing to an increase in cardiac fibrosis, the progression of ventricular dilatation and a decline in cardiac function. In turn, the haemodynamic overload, oxidative stress and endothelial dysfunction seen in dilated cardiomyopathy lead to an overwhelming inflammatory response for tissue repair, thereby creating a ‘vicious cycle’. BCR, B cell receptor; CD40L, CD40 ligand; MHC II, major histocompatibility complex class II; TCR, T cell receptor; TLR, Toll-like receptor.

The evidence for the association between genetic background and inflammatory response in DCM is limited but has been reported for other forms of cardiomyopathy. For example, preliminary evidence from autopsies of deceased patients who had ACM suggests overlapping infiltrates between classic myocarditis and genetic cardiomyopathy¹³⁶. Moreover, a myocarditis-like presentation with active myocardial inflammation has been reported in patients with ACM (genetic aetiology in 77% of patients), in the so-called hot phase of the disease¹³⁷. Genetic variants associated with DCM or ACM have been found in patients with acute and chronic myocarditis^138,139.

The detection of myocardial inflammation in DCM could provide an indication for immunomodulatory strategies. However, as discussed above, the effect of immunosuppressive therapy on LVRR, arrhythmias and survival remains unclear⁷⁰. Further evidence from multicentre trials is needed before anti-inflammatory treatment can be incorporated into guidelines for the management of patients with DCM.

Gene–environment interactions

Emerging evidence suggests a model whereby genetic predisposition to DCM can mediate different responses depending on environmental factors and, reciprocally, that environmental exposure might influence gene expression, supporting the two-hit hypothesis^140,141. Specifically, a gene–environment interaction between risk factors and TTNtv has been implicated in acquired DCM. First, observations of familial clustering of PPCM¹⁴², as well as co-occurrence with DCM^143,144, were confirmed in a genetic study in 172 women with PPCM¹⁴², in which the prevalence of truncating variants was 15%, with two-thirds located in the TTN gene. Second, carriers of TTNtv who consumed alcohol excessively (>21 units per week for men and >14 units per week for women) had a lower baseline LVEF than non-carriers with DCM who did not drink alcohol excessively¹⁴⁵. Third, a genetic relationship between DCM, susceptibility to cancer therapy-induced cardiomyopathy and adverse outcomes has been established, with pathogenic DCM variants (mainly TTNtv) found in 12% of patients with cancer therapy-induced cardiomyopathy¹⁴⁶. Adult patients with cancer therapy-induced cardiomyopathy and TTNtv had more HF-related hospitalizations and atrial fibrillation episodes and had worse cardiac function than those who did not carry TTNtv¹⁴⁶. Furthermore, the high prevalence of TTNtv in older patients (aged >60 years)³² suggests that the occurrence of multiple comorbidities, disease modifiers or ageing itself could have a central role in the penetrance and expressivity of a clinically overt DCM phenotype.

Very few environmental factors (alcohol, chemotherapeutic agents, pregnancy and age) have been investigated in the context of DCM, and even less is known about the effect of other modifiable second hits, such as physical exercise^147,148 or labile arterial hypertension¹⁴⁹. Moreover, the possible synergy of these factors in the manifestation of genetically determined cardiomyopathy has yet to be studied¹⁵⁰, and the small cohort sizes in existing studies limit the availability of adjusted data for potential contributory effects. Notably, the prevalence of TTNtv seems to be 1–3% in the general population^14,151,152, a percentage far above the prevalence of DCM (approximately 0.40%)², suggesting that most TTNtv carriers do not develop DCM. Furthermore, >90% of individuals with TTNtv do not develop PPCM, alcoholic cardiomyopathy or cancer therapy-induced cardiomyopathy¹⁵³, just as healthy individuals can have multiple pregnancies, consume excessive amounts of alcohol or be exposed to anthracycline-based agents without developing cardiomyopathy. These findings indicate that additional environmental, epigenetic or genetic factors might be involved. In this setting, variants of uncertain significance, variants within genes of uncertain association with DCM and common genetic variants could have a role in the variability of DCM. Early studies of polygenic risk scores established how common genetic variants associated with LV structure and function contribute to DCM risk among carriers of DCM rare variants¹⁵⁴. The findings tend to confirm the possibility that DCM might be caused by a high burden of common variants, evolving from an archetypal Mendelian disease to a polygenic disease^154,155 (Fig. 1).

Precision medicine

Current therapeutic strategies for DCM are based on guideline-directed medical therapies for HF^156,157 and have been previously reviewed^141,156,157. However, new approaches have shown promising results as discussed below, including risk stratification tools for precise identification of patients at risk of MACE and research efforts to develop agents that target and correct underlying molecular or genetic defects.

Risk stratification tools

Improved DCM phenotyping, large-scale genotyping and multicentre collaborative studies in large cohorts of patients with DCM have improved our understanding of disease outcome, enhanced prognostic stratification for HF and SCD, and have led to the development of risk stratification tools for use in clinical practice in distinct subgroups of patients with DCM.

The LMNA risk VTA Calculator SCD risk score for use in patients with LMNA variant cardiomyopathy was developed by analysing outcomes in 839 adult patients to improve the selection of candidates for an ICD²⁸. Furthermore, the PLN 5-year VA Risk Calculator was developed to identify candidates for ICD implantation by studying 679 carriers of the p.Arg14del founder variant in PLN, and provides a 5-year percentage risk of life-threatening ventricular tachyarrhythmias¹⁵⁸. In patients with ACM, which frequently overlaps with arrhythmogenic DCM, the ARVC Risk Calculator was developed to estimate the risk of ventricular arrhythmias within 5 years. This calculator has been shown to work well in those with variants in the PKP2 gene, but less well in those with DSP variants^159-161. For this reason, a large international DSP gene registry has been established, and the results of gene-specific risk-stratification tools are pending. FLNC variants also cause an arrhythmogenic form of DCM with a high risk of ventricular arrhythmias and SCD¹⁶². A large, international FLNC gene registry has also been established, and an FLNC variant-specific risk stratification tool is being developed.

LMNA variant cardiomyopathy carries a greater risk of progressive refractory HF than other genetic causes of DCM^10,49. In carriers of LMNA variants with DCM and limited therapeutic options, systolic function significantly improves with cardiac resynchronization therapy and is associated with increased survival¹⁶³. Improving phenotype characterization and prognostic stratification in large cohorts is crucial for the development of novel therapies. To this end, several collaborative efforts are ongoing, including the DCM SHaRe registry, a multicentre, international repository of clinical and genetic data on a large cohort of patients with DCM and their families (currently >6,000 patients). The aim of this project is to investigate disease outcomes and pave the way for future trials and biomarker discovery.

Targeted molecular therapy

A greater understanding of the molecular mechanisms and altered pathways leading to DCM has led to the discovery of novel therapies, and the myosin modulators are perhaps the most remarkable. In hypertrophic cardiomyopathy (HCM), mavacamten functions by decreasing the maximal actin-activated myosin ATPase activity, stabilizing myosin in a super-relaxed state and decreasing hypercontractility of the cardiac muscle. The drug has proven to be effective in the EXPLORER-HCM trial¹⁶⁴ and is currently approved by both the FDA and the European Medicines Agency for use in patients with obstructive HCM. Mavacamten has been also evaluated for the treatment of non-obstructive HCM in the MAVERICK-HCM trial¹⁶⁵. The use of myosin modulators has also been investigated in the treatment of patients with HF, including those with DCM, with the aim of activating the thin filament to induce an inotropic effect. Unfortunately, the findings from these investigations to date are less striking than the data for HCM. Another myosin modulator, omecamtiv mecarbil, showed a modest benefit among patients with HF with reduced ejection fraction in the GALACTIC-HF trial¹⁶⁶, but did not receive FDA approval because of an unfavourable risk-to-benefit ratio. Other myosin modulators, including danicamtiv, are under investigation.

In LMNA variant cardiomyopathy, phase I and II studies have shown that inhibition of p38 mitogen-activated protein kinase with ARRY-371797 could improve symptoms and signs of cardiac dysfunction^167,168. However, in the phase III, placebo-controlled REALM-DCM study¹⁶⁹, the results were less favourable and the trial was terminated by the sponsor due to futility. Lack of evidence of efficacy might be attributable to the complex functions of lamin A/C, the various cell types affected and the downstream effect of the active molecule.

In vitro experiments in human induced pluripotent stem cell-derived cardiomyocytes (iPSC-CMs) with FLNC-truncating variants (FLNCtv) have shown that delocalization of connexin 43 from the cell membrane causes nuclear translocation and subsequent activation of the platelet-derived growth factor receptor-α (PDGFRA) pathway, which results in arrhythmias¹⁷⁰. These changes were confirmed in explanted hearts from patients with FLNCtv. Treatment with the PDG-FRA inhibitor crenolanib improved contractile function and arrhythmic profile of FLNCtv iPSC-CMs¹⁷⁰, suggesting that pharmacological inhibition of the PDGFRA pathway could be a therapeutic strategy for DCM caused by FLNCtv.

Gene therapy

The ultimate therapy for DCM, when the pathogenic gene variant is identified, is the precise correction of the genetic defect. This approach has been explored through gene therapy studies in patients with HF for >10 years but has not been successful. However, new-generation gene therapy for monogenic inherited cardiomyopathies, including DCM, seems to be more promising¹⁷¹. Advances in this field include novel, high-efficacy vectors; novel capsids of adeno-associated virus (AAV) vectors with reduced immunogenicity; alternative delivery systems, such as extracellular vesicles (exosomes), which can be produced from autologous iPSC-CMs¹⁷²; and the use of modified RNA and circular RNA as therapeutic tools^173,174. Several potential problems with these approaches have been identified. First, the choice of the delivery system. Systemic delivery requires high doses, whereas intracoronary delivery is invasive but requires much lower doses and, therefore, has a better safety profile. Second, only genes <5 kb in size can be inserted into the AAV. Third, viral vectors have been associated with several safety concerns, including hepatotoxicity, acquired haemolytic uraemic syndrome, neurotoxicity, myocarditis and, potentially, oncogenicity¹⁷¹.

Trials of gene therapy that are ongoing in patients with DCM or overlapping cardiomyopathies are listed in Table 2. Three strategies are used for gene therapy: gene replacement therapy to correct the reduced functional protein, gene silencing to prevent the expression of the mutant protein, and direct genome editing⁴. An early gene replacement therapy for a monogenic cardiomyopathy has been studied in patients with Danon disease, an X-linked cardiomyopathy caused by lack of lysosome-associated membrane glycoprotein 2 (ref. 175). RP-A501 uses a recombinant cardiotropic AAV9 enclosing LAMP2 isoform B. The phase I trial showed that RP-A501 has a good safety profile, is associated with cardiomyocyte histological correction, and leads to improvement or stabilization of clinical status and a reduction in LV hypertrophy and levels of cardiac biomarkers (unpublished data). These findings paved the way for the ongoing phase II trial¹⁷⁵. Trials on gene replacement therapy are also ongoing for other DCM-associated genes, including BAG3 (AVB-401 in phase 0) and PKP2 (TN-401 in phases 0 and Ib). A gene-silencing approach has been developed for LMNA variant cardiomyopathy using short hairpin RNA (PHL-001) to achieve allele-specific silencing of mutant LMNA, and is currently in preclinical studies.

Table 2 ∣.

Gene therapy trials in patients with DCM or overlapping cardiomyopathies

Disease	Gene	Treatment	Mechanisms	Developer	Phase	Status	ClinicalTrials.gov IDa
Danon	LAMP2	RP-A501	AAV9.LAMP2B	Rocket Pharmaceuticals	Phase I	Active, not recruiting	NCT03882437
Danon	LAMP2	RP-A501	AAV9.LAMP2B	Rocket Pharmaceuticals	Phase II	Recruiting	NCT06092034
DCM, BAG3 LoF	BAG3	AVB-401	AAVrh74.BAG3	Solid Biosciences	Preclinical	NA	NA
DCM, BAG3 LoF	BAG3	Registry	AAV9.BAG3	Pfizer, AstraZeneca, Axion	Phase 0 registry	Active, not recruiting	NCT05981092
DCM, BAG3 LoF	BAG3	REN-001	AAV9.BAG3	Renovacor, Rocket Pharmaceuticals	Preclinical	NA	NA
ACM, PKP2 LoF	PKP2	TN-401	AAV9.PKP2	Tenaya Therapeutics	Phase 0 registry	Recruiting	NA
ACM, PKP2 LoF	PKP2	TN-401	AAV9.PKP2	Tenaya Therapeutics	Phase Ib	Pending	IND application
ACM, PKP2 LoF	PKP2	RP-A601	AAVrh74.PKP2	Rocket Pharmaceuticals	Phase I	Recruiting	NCT05885412
ACM, PKP2 LoF	PKP2	LX2020	AAVrh.10hPKP2	Lexeo Therapeutics	Phase I and phase II	Recruiting	NCT06109181
ACM, PKP2 LoF	PKP2	BMN365	AAV.PKP2	BioMarin Pharmaceuticals	Preclinical	NA	NA
DCM	LMNA	PHL-001	Short hairpin RNAs for mutant allele-specific gene silencing	Phlox Therapeutics	Preclinical	NA	NA
HCMb	MYBPC3	TN-201	AAV9.MYBPC3	Tenaya Therapeutics	Phase Ib	Recruiting	NCT05112237
HCMb	MYBPC3	BMN293	AAV9.MYBPC3	BioMarin Pharmaceuticals, Dinaqor	NA	NA	NA
Non-obstructive HCMb	NA	HTX-001	Antisense oligonucleotide targeting Wisper long non-coding RNA	Haya Therapeutics	Preclinical	NA	IND enabling

AAV, adeno-associated virus; ACM, arrhythmogenic cardiomyopathy; BAG3, BAG family molecular chaperone regulator 3; DCM, dilated cardiomyopathy; HCM, hypertrophic cardiomyopathy; IND, investigational new drug; LoF, loss of function; NA, not available; PKP2, plakophilin 2. ^a ClinicalTrials.gov accessed on 4 September 2024. ^b Sarcomeric HCM included for comparison.

The ultimate, but also most challenging, approach is correction ofthe variant, which might be accomplished by genome editing using CRISPR–Cas9 technology. This strategy is already in human trials for haematological disorders, such as β-thalassaemia and sickle cell disease, and for cancer. In DCM, promising investigations in mice have shown feasibility and prevention of disease onset in cardiomyopathies associated with MYH7 (refs. 176,177) or RBM20 (ref. 178) variants. Although precision genome editing holds great promise, human trials should be conducted with caution due to safety concerns related to possible off-target effects and germline genome editing¹⁷⁹. Furthermore, advances in artificial intelligence and machine learning offer the ability to establish the phenotypic heterogeneity underlying DCM and to apply the huge volume of clinical data to guide patient management^3,180.

Conclusions

DCM is a complex disease with various underlying pathophysiological mechanisms. Improvements in knowledge and technology have enabled the identification of genetic aetiology in approximately 40% of patients¹⁰. However, non-genetic factors are also important, and a growing body of evidence links genetic background with concomitant non-genetic triggers or precipitating factors (the two-hit hypothesis). A better understanding of aetiology and pathogenetic mechanisms of DCM are paving the way for future therapies, including small molecules, RNA and gene therapy, and measures for the prevention of arrhythmic death. Increasing access to population-based biobanks and the creation of large multicentre registries will help researchers and clinicians fill the gaps in our understanding of the pathogenesis of DCM and the risk and prognosis in patients with DCM, and facilitate the development of novel precision therapies.

Key points.

• Dilated cardiomyopathy (DCM) is a heterogeneous disease with multiple causes and high variability in phenotype presentation and outcomes.

• Genetic testing has been recommended in guidelines as a fundamental step in the clinical decision-making process; the results can guide not only family screening but also aetiological characterization and risk stratification.

• Genotype–phenotype correlations can aid clinicians in prioritizing genetic testing in patients with a higher likelihood of identifying a disease-causing variant, particularly for genes associated with major arrhythmic events.

• A strong interaction exists between the genetic background and environmental exposures, which can lead to different phenotypic manifestations of the disease (that is, the two-hit hypothesis).

• Advances in the management of DCM are moving towards a precision medicine approach to the diagnostic work-up and treatment.

• Future research and resource allocation will focus on the identification of disease-modifying treatments for DCM, which include molecular targeted and gene therapies.

Related links

ARVC Risk Calculator: https://arvcrisk.com/

ClinGen database: https://www.clinicalgenome.org/

DCM SHaRe registry: https://www.theshareregistry.org/

LMNA risk VTA Calculator: https://lmna-risk-vta.fr/

PLN 5-year VA Risk Calculator: https://plnriskcalculator.shinyapps.io/final_shiny/