CHAPTER 58: DILATED CARDIOMYOPATHY

Dilated cardiomyopathy (DCM) is a chronic heart muscle disease characterized by “the presence of dilatation and systolic impairment of the left or both ventricles unexplained by abnormal loading conditions or coronary artery disease sufficient to cause the observed myocardial dilation and dysfunction.”¹ According to the American Heart Association (AHA) classification, “dilated forms of cardiomyopathy are characterized by ventricular chamber enlargement and systolic dysfunction with normal left ventricular (LV) wall thickness.”²

In this chapter, DCM is intended as a primary disease of heart muscle. More than 60% of DCMs are proven to be familial diseases with identifiable genetic defects.^3,4 Acquired disorders manifesting with the DCM phenotype, or DCM phenocopies (defined as environmentally induced phenotypes mimicking one usually produced by a specific genotype), are sporadic and are categorized as nongenetic DCM. This distinction is essential to separate genetic DCM from acquired, nongenetic, and potentially reversible phenotypes arising from protean causes. The therapeutic options for acquired DCM often differ from those for familial DCM (FDCM). Ischemic heart disease may manifest as, or evolve into, a DCM-like phenotype; it is discussed further in Chap. 43.

In its descriptive definition, DCM represents the end phenotype of heart muscle damage induced by different causes; the disease mechanisms are increasingly being identified by the implementation of molecular and genetic assays for patients and families, high-resolution and functional imaging, and novel biomarkers. In this end stage, most DCMs look phenotypically alike (Fig. 58–1). Intermediate phenotypes are manifest by borderline, persistent LV dilation and/or dysfunction and may present with arrhythmias and/or conduction disease, now recognized as “early DCM.”⁵ In genotyped families, the preclinical phase of the disease can be diagnosed in phenotypically healthy, but genetically affected, relatives of probands with DCM. Therefore, DCM can be grouped mechanistically as genetic and nongenetic.

Current estimates of incidence are derived from phenotype-based studies. Phenotype-based incidence is estimated as 6.0 per 100,000 person-years and prevalence is 36.5 per 100,000 person-years (about 1:2500).⁶ These epidemiology data are based on clinical diagnoses and have been obtained in the pregenetic era. New estimates are expected in the short term,⁷ based on data obtained with early diagnoses, systematic family screening and monitoring, genetic testing, and advanced diagnostics for nongenetic DCM.^{8,9,10,11,12,13,14,15,16} As long-term follow-up of FDCM increases, the overall prevalence of disease is anticipated to increase.

When DCM is defined by etiology, the prevalence data of each specific disease manifestation with a DCM phenotype is far less common, placing most DCM subtypes in the context of rare diseases (< 1:2000). Therefore, the burden of DCM in general, and of FDCM in particular, is a dynamic issue that should be regarded as evolving and deserving of further efforts to generate a precise epidemiology of DCM.

DCM typically manifests with systolic heart failure (HF). The spectrum of symptoms is wide and nonspecific; it includes fatigue, breathlessness, palpitations, and signs of congestion/fluid retention in the presence of structural cardiac abnormalities corresponding to the diagnostic definition.^1,17 The clinical diagnosis is made in the presence of LV systolic dysfunction (impaired LV ejection fraction measured using two-dimensional echocardiography) and LV dilatation (as defined by Z-score > 2 standard deviations for LV end-diastolic volumes or diameters, according to nomograms corrected for body surface area, height, and age). Coronary angiography is routinely used to exclude coronary atherosclerosis.^1,2,17 Coronary computed tomography (CT) angiography may be considered as an alternative anatomic assessment for the evaluation of obstructive coronary artery disease. The presence of flow-limiting luminal stenosis in one or more epicardial coronary arteries is considered sufficient to assign the cause of DCM to coronary artery disease. However, coronary and/or systemic atherosclerosis and DCM can exist concurrently and be demonstrated in genetic DCM.¹⁸ Individualized decision making and testing are often necessary to clarify the extent of the contribution made by coronary artery disease in the pathogenesis of DCM. Functional testing, as with myocardial perfusion imaging, may determine presence and/or extent of ischemia and may be useful in this regard.

In the past decade, the clinical use of cardiac magnetic resonance (CMR) has entered routine evaluation, both helping to confirm the diagnosis and adding information on the type and extent of myocardial fibrosis.^19,20,21 CMR may further reveal hypertrabeculation or LV noncompaction (LVNC) patterns that can be associated with DCM (see Chap. 60). Three-dimensional echocardiography may add information on valvular anatomy and remodeling morphology, particularly in patients with limited views on two-dimensional echocardiography.²² Contrast echocardiography enhances endocardial border definition, increases accuracy to two-dimensional and three-dimensional volume assessment, and highlights trabecular anatomy; it is not needed in routine evaluation, but may help in cases of suboptimal image quality.²³ Myocardial deformation imaging techniques²⁴ and myocardial tagging by CMR are promising contributors to the diagnosis of “early” phenotypes.^5,25

In more than 60% of cases of DCM, the disease is familial, as proven by clinical family screening demonstrating that more than one member is either affected or show traits that predict the development of the disease.^5,26 The diagnosis of FDCM is formulated when at least two members of the same family are affected. The clinical workup is a three-step investigation:

1. In the proband, deep phenotyping should include a general medical evaluation of anthropometric features and the integument, hair, vision, and musculoskeletal, auditory, and nervous systems.²⁷ General phenotype examination can add information about clinical traits potentially associated with the DCM and provides robust clinical data for geno-phenotype correlations and segregation studies in families.

2. The next step is genetic counseling with pedigree construction and information provided to patients about why family studies are necessary when a genetic disease is suspected. The collection of the clinical history of relatives and tracing of clinical reports of deceased relatives are uniquely useful for correct interpretation of fatal events in families.^26,27

3. The third step is family screening, which is independent of genetic testing. Clinical family screening is based on physical examination, electrocardiography (ECG), transthoracic echocardiography, and biochemical testing, followed by regular monitoring of relatives. Once completed, information gleaned from the clinical family screening may modify the original pedigree. Determination of the type of inheritance of the DCM results from integration of pedigree and family data, and is essential for guiding and interpreting data generated by genetic testing.^28,29,30

Data achieved in family screening studies of consecutive patients demonstrate that relatives of patients with DCM can:

• Be affected, and therefore present the criteria for diagnosis, but be asymptomatic and unaware of their disease

• Manifest early, persistent (confirmed on more than one clinical occasion) signs of LV remodeling and/or borderline dysfunction⁵

• Demonstrate electrocardiographic abnormalities such as conduction disease, short PR interval/Wolff-Parkinson-White pattern, prolonged QT interval, T-wave abnormalities, early repolarization, or abnormal high or low voltages of the QRS^5,27

• Demonstrate abnormal biochemical markers such as increased serum creatinine phosphokinase or lactic acid

• Exhibit extracardiac traits that may also occur in probands, such as hearing loss, diabetes mellitus, cutaneous lesions, skeletal abnormalities, abnormal faces, or ocular defects²⁷.

Family screening may by itself offer added value for family health.^3,5,27,28,30 Monitoring of family members helps generate reliable data on the natural history of the disease (Fig. 58–2). DCM is often characterized by subtle onset and slow progression; it shows a typical age dependency of the phenotype. Overall, clinical family screening is integral and indispensable for the diagnosis of FDCM.

Inheritance and Phenotype Variability

The increasing identification of genetic defects segregating with a particular phenotype in families confirms the genetic origin of the majority of FDCMs²⁶ (Table 58–1). Mendelian laws and rules of matrilinear inheritance govern the transmission of cardiomyopathies (see Chap. 57). Most FDCMs are autosomal dominant diseases; a minority shows autosomal recessive, X-linked recessive, and matrilineal inheritance. X-linked dominant forms are exceptional.^28,30 In autosomal dominant FDCM, each patient has a 50% probability of transmitting the mutation to offspring; in autosomal recessive FDCM, each parent is a healthy carrier of the mutation: the risk of transmission of the mutated allele from both parents is 25%. In X-linked recessive FDCM, the risk is 50% when the carrier is the mother; when the proband is the father, all daughters are obligate carriers, whereas all sons are noncarriers. In families with matrilineal cardiomyopathies, females transmit the mutation to both male and female offspring; accordingly, both genders can be affected. Vice versa, male carriers do not transmit the mutation. Mitochondrial DNA-related cardiomyopathies can demonstrate both dilated and hypertrophic phenotypes, with the latter frequently evolving through dilation and systolic dysfunction, mimicking typical DCM in later stages.²⁹

Clinical traits, both cardiac and extracardiac, may help formulate pretest hypotheses.²⁷ ECG or imaging markers, skeletal muscle disease, auditory defects, ocular abnormalities, gastrointestinal disturbances, renal disease, cutaneous lesions, cognitive impairment, and cryptogenic stroke can be associated with DCM caused by different disease genes. Cardiomyopathies caused by mutations in mitochondrial DNA genes are commonly observed in the clinical context of syndromes with involvement of skeletal muscle (myopathies), brain (cryptogenic stroke), ear (sensorineural hearing loss), and kidney (renal failure); in mitochondrial diseases, diabetes is common and should be considered a phenotypic trait contributing to characterization of the phenotype in the family^26,27,28 (Table 58–2).

Genetic Heterogeneity

Genetic DCM is the paradigmatic example of how “similar phenotypes” may be caused by “different genes and mutations.” More than 100 genes are now included in the list of disease and candidate genes (see Table 58–1). Attempts to group disease genes according to the pathways or structures in which their products are involved may outline macro-groups of disease genes^{31,32,33,34,35,36,37,38,39}:

• Sarcomere genes (TTN, MYH7, MYBPC3, TNNT2, TNNI3, MYL2, MYL3) in 25% to 30% of DCM patients.³¹ Although truncation predicting mutations in TTN have been reported in 25% of DCM patients,³² confirmatory data are necessary, because 1 in 500 individuals carried truncation-predicting variants in control series³³

• Nuclear envelop genes (LMNA, EMD, SYNE1, TMPO) in about 7% to 10% of cases, with LMNA mutations accounting for the majority of DCM in this subgroup^34,35,36,37

• Z-disc genes (ACTN2, ANKRD1, BAG3, CRYAB, CSRP3, FHL2, LDB3, MYPN, MURC, NEXN, PGM1, VCL), all together accounting for less than 10% of DCM^38,39

• Genes coding intermediate filaments connecting sarcolemma with sarcomeres (DES) in about 1% of cases⁴⁰

• Mitochondrial genes, both nuclear (eg, G4.5, CTF1, SDHA, DNAJC19) and mitochondrial DNA genes, in less than 5% of cases⁴¹

• Dystrophin defects in 6% to 7% of DCM in consecutive male series and dystrophin-associated glycoprotein complex (ie, SGCA, SGCB, SGCD) in less than 1% of DCM cases^42,43,44

• Ion channels (SCN5A)^45,46 or genes such as ABCC9 or CHRM2, in less than 1% of cases^47,48

• Genes whose products are active in Golgi apparatus machinery (eg, FKTN) or in the sarcoplasmic reticulum (PLN) in less than 1% of cases.⁴⁹ The prevalence of PLN-related DCM varies in different geographic areas. Founder mutations such as PLN R14del in the Netherlands accounts for more than 10% of all DCM among the Dutch population.⁵⁰

Mechanisms of Myocardial Damage

The mechanisms through which mutations in different genes cause functional and structural myocyte damage depend on the gene and the role of its product(s) in the myocyte, type of mutation and its effects on the protein expression, epistasis (the action of one gene upon another), and epigenetic factors. The genetic heterogeneity of DCM indicates that different types of molecular insults may end in similar phenotypes.⁵¹ Mechanisms such as haploinsufficiency (a single functional copy of a gene is insufficient to maintain normal function) have been demonstrated in dilated cardiolaminopathies and dilated cardiodystrophinopathies.^52,53,54 Two genes (LMNA and DMD) typically cause DCM, with or without arrhythmias, but do not cause hypertrophic cardiomyopathy (HCM). However, most DCM genes also cause other types of cardiomyopathy (HCM, restrictive cardiomyopathy, and arrhythmogenic right ventricular cardiomyopathy). The paradigmatic example is DCM caused by mutations in sarcomeric genes such as MYH7; the understanding of the mechanisms that cause either HCM or DCM in carriers of MYH7 mutations is still far from being elucidated. It has been proposed that mutant myosins with enhanced contractility lead to HCM, whereas those displaying decreased contractility lead to DCM. Gain or loss of function could be the primary consequence of a specific mutation.⁵⁵

Genetic Testing

Genetic testing is now routinely performed using massive parallel sequencing of multigene panels (next-generation sequencing, NGS) for both nuclear genes^3,4 and mitochondrial DNA.⁵⁶ The time needed for testing is short, and the costs are lower than with Sanger-based technologies. The mismatch between fast rate of detection and the time needed for developing functional studies that assess the effects of mutations explains the multiple pipelines generated in the past decade to contribute to the interpretation of novel genetic variants including those with uncertain significance (see Chap. 9).

For several disease genes, a proven cause-and-effect relationship between gene mutations and disease is now robust enough to assign a diagnostic role to genetic testing; the causal link of the mutation with the disease is still provisional for several other candidate genes, especially those reported in one or a few families and still unconfirmed in larger series. Interpretation of genetic testing is the challenge of the next decade, considering that when multiple genes are tested and the gene panels include giant genes such as titin (TTN) or large and complex genes such as Duchenne muscle dystrophy (DMD) and obscurin, the probability of finding more than one mutation per individual increases.³ Further developments are needed to generate efficient tools for exploring the functional effects and pathologic role of mutations, especially when considering that most heritable DCMs are Mendelian diseases transmitted through an autosomal dominant mode, a rule that does not change even in families in which more than one mutation is identified. This implies that one major (or leading) mutation or two mutations inherited from the same affected parent plays a causative role, with other variants potentially acting as phenotype modifiers. Misinterpretation of the role of the mutation in the pathogenesis of the DCM may have severe implications for the patient and family, especially when procreative plans include prenatal diagnosis.

The proportion of DCM causally linked with known genes can be posted as definite (ie, LMNA, DYS, and sarcomere genes) or considered provisional, especially among genes that have been reported in unique families/cases. Ethnic variations and founder mutations may explain different prevalence of mutations in disease genes; for example, PLN p.(Arg14Del) is rare in the general population but causes more than 10% of all FDCMs in the Netherlands.⁵⁰ Most studies available at the end of 2015 have been performed either testing a single gene or a few genes in different clinical series. In the future, the genetic epidemiology of DCM could change on the basis of the results achieved with NGS testing panels of genes in large clinical series. At present, genetic defects are identified in about 60% of FDCM; this implies that additional disease genes will be added to this list.

The interpretation of the role of a putative mutation in a given family is a multistep process that is:

1. Predicted by the results and analysis of NGS tests: Pipelines starting from raw data that include thousands of genetic variants/patients and ending with one or a few probable or possible disease mutation(s).

2. Based on consolidated evidence of pathogenicity for known mutations: automated collection of information on mutations that have been previously reported and proven as pathogenic requires disease specific expertise and regular monitoring of disease databases. Identification of a mutation, while integral to the assignment of pathogenicity, cannot by itself, at least in some instances, be sufficient to establish pathogenesis. For example, although truncation-predicting mutations are likely to be pathologic, a few of them occur in nonaffected individuals and should be considered either nonpathologic or variants of uncertain significance.

3. Confirmed by in vivo or in vitro functional and pathologic studies (Fig. 58–3): either endomyocardial biopsy with immunohistochemical study of the protein coded by the mutated gene or in vitro studies on fibroblasts or induced pluripotent stem cell–derived myocytes obtained from circulating cells or fibroblasts from patients who carry the putative mutation.

4. Integrated into a clinical context, based on segregation studies in families (Fig. 58–4): mutated family members are affected and nonmutated members are not affected. Healthy mutation carriers are usually younger than affected mutated family members.

The overall evidence gained by the combination of the above criteria should establish whether a mutation is pathologic and, therefore, should be posted as a disease-causing mutation.

Grouping Genetic Dilated Cardiomyopathy by Disease Gene

Although to date more than 100 disease and candidate genes have been associated with DCM, only a few of them have been confirmed and replicated in more clinical series and families. In addition, clinical markers typically recurring in DCM caused by mutations in these genes are now consolidated contributors in the pretest clinical orientation about the possible disease gene to be tested in the probands.²⁷ The following examples have been selected based on different criteria: cardiolaminopathies because of their high arrhythmogenic risk (they represent the first genetic DCM formally recognized based on the disease gene and types of mutations in guidelines for primary prevention of sudden death with implantable cardioverter-defibrillator [ICD] implantation); cardioemerinopathies because, although very rare, their phenotype may look similar to cardiolaminopathies (DCM with conduction disease and myopathy) but with different pattern of inheritance (X-linked recessive vs autosomal dominant); cardiodystrophinopathies because they may represent the major or unique phenotype in male patients with DCM and are associated with increased serum creatine phosphokinase in > 80% of cases; cardiozaspopathies because of the frequent association with prominent trabecular anatomy of the left ventricle; cardiomyosinopathies because defects of the structural and regulatory myosin complex of the sarcomere may cause typical HCM, but also DCM and HCM-DCM (ie, a phenotype that in the end-stage phases is similar to DCM but frequently maintains electrocardiographic features suggestive of LV hypertrophy); dilated cardioMITOmyopathies because of their high malignancy and the syndromic context in which the cardiomyopathy is observed; cardiotitinopathies because of their high prevalence in DCM series; and phospholambanopathies because of the high arrhythmogenic risk and peculiar ECG markers.

Dilated Cardiolaminopathies

This subgroup of DCM can be described as dilated cardiolaminopathies, or DCM-LMNA; an individual description can be provided by the simple use of the MOGE(S) system (morphofunctional phenotype, organ involvement, genetic inheritance pattern, etiologic annotation including genetic defect or underlying disease/substrate, and functional status of the disease) supported by on online tool at http://moges.biomeris.com/moges.html (see Chap. 57). DCM-LMNA is phenotypically characterized by the association of the DCM with conduction system disease, which occurs in up to 80% of patients with cardiolaminopathies.⁵⁷ In the past 15 years, the association between LMNA gene mutations and DCM with conduction system disease has been replicated and confirmed.^58,59,60 Dilated cardiolaminopathies constitute 6% to 7% of all DCMs in consecutive series.³⁹ The development of conduction system disease usually precedes the appearance of the DCM^7,16; the natural history is characterized by a long asymptomatic phase in which regular monitoring demonstrates a progressive prolongation of the PR interval and slowly progressive LV dilation and/or dysfunction. Cardiolaminopathies display high arrhythmogenic potential for both atrial and ventricular arrhythmias. The high risk of life-threatening ventricular arrhythmias is one of the characteristics of cardiolaminopathies, and such arrhythmias may manifest even in mildly dilated and dysfunctioning hearts.^59,60,61,62 Recent guidelines on the primary prevention of sudden cardiac death (SCD) recommend ICD implantation in patients with DCM and a confirmed disease-causing LMNA mutation and clinical risk factors (Class IIa, Level B)⁶³ such as nonsustained ventricular tachycardia during ambulatory ECG monitoring, LV ejection fraction (LVEF) < 45% on initial evaluation, male gender, and nonmissense mutations (insertion, deletion, truncation, or mutations affecting splicing). These factors correspond to those previously identified in a large European cohort of patients with dilated cardiolaminopathy.⁶⁰

Dilated Cardioemerinopathies

Dilated cardioemerinopathies, or DCM-EMD, constitute a small subgroup of DCM but also provide a paradigmatic example of rare DCM that may phenotypically mimic cardiolaminopathies.^30,35 They are, in fact, typically associated with conduction disease; different from cardiolaminopathies, they are inherited as X-linked recessive diseases and associated with muscle dystrophy. This latter feature is also possible in DCM-LMNA, but less common. Emerin is a ubiquitous protein, major component of the nuclear envelope, and member of the nuclear lamina-associated protein family. Defects in emerin (EMD) cause X-linked recessive Emery-Dreifuss muscular dystrophy. Cardiac involvement is common and manifests with palpitations, presyncope and syncope, poor exercise tolerance, and HF. The clinical management includes treatments for cardiac arrhythmias, atrioventricular conduction disorders, and congestive HF, including antiarrhythmic drugs, cardiac pacemaker, ICD, and heart transplantation for end-stage HF.^35,64 Conduction disease is, therefore, the marker of the disease in patients with X-linked inheritance and DCM; the immediate clinical differential diagnosis with autosomal dominant laminopathies is the inheritance pattern.

Dilated Cardiodystrophinopathies

Dilated cardiodystrophinopathies, or DCM-DYS, typically manifest in male patients who carry large DMD gene rearrangements, both in-frame and out-of-frame deletions (> 80%)^42,43; in a minority of cases, mutations are either nonsense or missense or splice. Mutations are associated with loss of protein expression at the level of the cardiomyocyte sarcolemma.⁴³ DMD gene (MIM#300377) mutations cause the following three phenotypes: Duchenne muscular dystrophy (MIM#310200), Becker muscular dystrophy (MIM#300376) and X-linked DCM (MIM#302045). Most data on cardio-phenotypes have been generated in series of patients with Duchenne muscular dystrophy and Becker muscular dystrophy; single case reports and a few series explored patients presenting with DCM phenotype at onset.^42,43 Milasin et al⁶⁵ first described a family with X-linked DCM associated with a splice donor site mutation in the DMD gene. Antidystrophin immunostain showed reduced, but normally distributed, protein in the skeletal muscle and no detectable protein in the myocardium, suggesting that selective heart involvement in DCM-DMD is related to the absence in the heart of a compensatory expression of dystrophin from alternative promoters.⁶⁵ This case generated a cardio-specific interest; further reports of single cases demonstrated that mutations in the DMD gene may exclusively or predominantly affect the heart, with normal or early involved skeletal muscle.^66,67 The precise diagnosis can be missed and eventually determined after heart transplant, when muscle dystrophy may manifest and be attributed to statin treatments.⁶⁸

Although rare, the above past studies facilitated the feasibility of heart transplantation in either patients with X-linked DCM or Becker muscular dystrophy with DCM as a major and initial presentation of dystrophinopathy (see Fig. 58–3). The diagnosis of X-linked DCM-DYS is feasible by endomyocardial biopsy, which demonstrates loss of protein expression; genetic testing confirms the diagnosis.^42,43 X-linked DCM-DYS carries a low arrhythmogenic risk even when patients show severely dilated and dysfunctioning hearts, requiring ICD for primary prevention of SCD according to current guidelines. No ICD shock was observed during a median follow-up of 14 months (interquartile range, 5-25 months) in 34 patients with DCM caused by defects of dystrophin.⁴³ DCM-DMD may show LVNC, which occurs in about 30% of cases and is now proposed as a prognostic marker.^69,70 The recent introduction of ventricular assist device implantation in children affected by DMD opens novel avenues for treatment of the most severe forms of DCM-DMD.^71,72

Dilated Cardiozaspopathies

This small subgroup of DCM is caused by mutations in the LDB3 gene that codes the Z-disc Cypher-ZASP protein. LDB3 mutations were originally reported in adult-onset, isolated, dilated LVNC cardiomyopathy⁷³ (OMIM*605906.0007) (see Chap. 60). Therefore, LVNC and LV hypertrabeculation are the clinical markers that characterize DCM hearts of patients who are carriers of mutations in LDB3.^74,75,76,77 Mutations in the LDB3 gene seem to also be associated with HCM⁷⁸ and myofibrillar myopathy type 4.⁷⁹ The more common “variant” in this gene is p.(D117N), which has been recently reported to represent a polymorphism in a Bedouin population in which it occurs in 5.2% of nonaffected individuals.⁸⁰ The possibility exists that mutations in LDB3 may influence trabecular anatomy, leading to either increased trabeculation or LVNC, which by itself only describes an anatomic feature and not necessarily a functional phenotype. Although available data are not sufficient to post LDB3 as a disease gene specifically associated with both DCM and LVNC, this latter marker should be a matter of further investigation in LDB3 hearts to definitely assess its role in the trabecular anatomy of DCM hearts.

Dilated Cardiomyosinopathies

Recent studies have focused on the question of why mutations in the MYH7 gene that typically cause HCM may also be associated with the development of DCM⁵⁵ rather than the most common HCM. In murine models investigating the S532P and F764L mutations in the MYH7 gene and associated with the DCM phenotype, the molecular mechanism leading to DCM may be the depressed molecular function in cardiac myosin, which may initiate heart remodeling and pathologic dilation. Accordingly, mutations that lead to DCM may be those that are associated with loss of function; those that are associated with gain of function may be associated with HCM.^55,81 The reconstitution of the entire contractile system (actin and myosin^82,83 and tropomyosin 1, troponin C, troponin I, and troponin T^84,85) seems to be necessary to fully understand the effects of MYH7 mutations in HCM and DCM.⁸⁶ Alternative explanations include the possibility of compound mutations in mitochondrial DNA genes and in the MYH7 gene; patients who carry the sole MYH7 mutation demonstrate HCM, whereas those carrying both MYH7 and mitochondrial DNA mutations develop hypertrophic dilated cardiomyopathy with an end phenotype resembling DCM.⁸⁷ Yet another explanation is the presence of more than one mutation in the same patient; depending on the mutation, the DCM phenotype should be explained by a complex genetic make-up when more than one mutation is present and contributing to the final phenotype.^28,30 Overall, the role of MYH7 in DCM remains a matter of research and deserves specific investigation for any novel mutation identified in patients with DCM.

Dilated CardioMITOmyopathies

In this subgroup of patients, the DCM is usually observed in the context of multiorgan syndromes with a peculiar risk of complications such as cryptogenic strokes and intolerance to several drugs.^{28,29,30,41,56} Mitochondrial cardiomyopathies can be caused by mutations both in mitochondrial DNA genes (maternal inheritance) and in nuclear genes (Mendelian inheritance: no male passes down the disease to children) coding mitochondrial proteins. They are characterized by either hypertrophic phenotype evolving through dilated and dysfunctioning hearts or DCM (Fig. 58–5). The clinical manifestations associated with mutations in mitochondrial DNA genes depend on the grade of heteroplasmy in affected organs. They are commonly observed in families in which mutation carriers also express noncardiac traits such as hearing loss, palpebral ptosis, myopathy, renal failure, cryptogenic stroke, diabetes, optic neuritis, and/or retinitis pigmentosa. Sequencing of mitochondrial DNA, either by Sanger-based techniques or NGS tools, identifies the causative mutation. Nuclear “mitochondrial” cardiomyopathies are inherited according to Mendelian rules: they can be autosomal dominant or recessive. Typical cardiac phenotypes are primarily DCM, such as in the autosomal recessive DCM caused by ANT1 mutations⁸⁸ or in autosomal dominant DCM and progressive external ophthalmoplegia caused by POLG1 mutations.⁸⁹

Dilated Cardiotitinopathies

The TTN gene encodes the largest human protein and the third most abundant striated muscle protein.⁹⁰ Of the different titin isoforms, the N2B and N2BA isoforms are predominantly expressed at the cardiac level⁹¹; the cardiac isoform Novex-3 titin (5600 amino acids) is less represented in cardiac tissue than full-length titin.⁹² Truncation-predicting mutations have been identified in 25% of FDCMs and in 18% of sporadic DCMs.³² The penetrance of TTN truncating mutations was more than 95% for patients who were older than 40 years of age. Clinical manifestations, morbidity, and mortality were similar to those observed in DCM patients who were noncarriers of TTN mutations. DCM was not accompanied by conduction system or skeletal muscle disease.³² A further study in women with peripartum cardiomyopathy (PPCM), showed a distribution of truncating variants remarkably similar to that found in patients with DCM.⁹³ Missense variants have been further investigated in DCM distinguishing “bioinformatically severe” TTN versus nonsevere variants; carriers of “severe TTN variants” showed a clinical course similar to that of noncarriers.⁹⁴ TTN mutations have been involved in HCM (MIM#613765); muscular dystrophy of the Limb-Girdle, type 2J (MIM#608807); autosomal recessive early-onset myopathy with fatal cardiomyopathy (MIM#611705); tardive tibial muscular dystrophy (MIM#600334); and proximal myopathy, with early respiratory muscle involvement (MIM#603689).^95,96,97,98 Because 1 in 500 individuals in the general population carries a truncation variant in the TTN A-band, the interpretation of such mutations in DCM should be cautious.³³ The relevance of defining the precise role of TTN in DCM is now supported by the possibility that antisense-mediated exon skipping can be explored as a therapeutic strategy.⁹⁹

Dilated Phopsholambanopathies

Dilated phopsholambanopathies are a distinct group of DCM caused by mutations in the phospholamban (PLN) gene that encodes a protein expressed in the sarcoplasmic reticulum membrane. Phospholamban inhibits cardiac muscle sarcoplasmic reticulum Ca²⁺-ATPase (SERCA2a) in the unphosphorylated state. Mutations in PLN impair sarcoplasmic reticulum calcium homeostasis and cause DCM.^50,100,101 Less commonly, mutations in PLN cause HCM.¹⁰² p.(Arg14del) is the most frequently identified mutation in Dutch cardiomyopathy patients (10%-15%), both DCM and arrhythmogenic phenotypes.⁵⁰ Patients with p.(Arg14del) diagnosed with DCM demonstrate high arrhythmogenic risk and SCD as first disease presentation.¹⁰³ A common ECG pattern in PLN mutation carriers is a low-voltage QRS complex and inverted T waves in leads V₄ to V₆.¹⁰⁴ In non-Dutch patients, PLN mutations are less common but equally malignant.^105,106 Therefore, although rare, PLN mutations can be clinically suspected in patients with DCM, high arrhythmogenic risk, and low voltage on ECG.

Future Developments in Genetic Dilated Cardiomyopathy

As the number of patients and families diagnosed with DCM caused by different disease genes increases, the possibility of generating large groups of patients with shared genotypes and phenotypes will provide data for both novel and precise genetic epidemiology and for large disease-specific databases. Ideally, precisely diagnosed genetic DCM should be grouped per disease genes and mutations causing similar effects. Genetic epidemiology will help highlight proportions of different genetic DCM, guiding assignment of clinical priorities and engaging all stakeholders in research and development programs for disease-specific treatments aimed at managing the basic genetic defects and not just the phenotype.

Nongenetic DCM includes etiologically heterogeneous diseases whose onset and progression may largely differ from those of genetic DCM. Theoretically, DCM phenotypes associated with acquired diseases/conditions and/or toxic exposures are potentially reversible with recognition, control, and/or removal of the offending cause. The major subgroups of acquired DCM include the following:

• Autoimmune/immune-mediated DCM may manifest in patients with systemic autoimmune disease such as systemic sclerosis, systemic lupus erythematosus, rheumatoid arthritis, Sjögren syndrome, and polymyositis/dermatomyositis (see Chap. 100). The Churg-Strauss syndrome (historically included in this group of diseases) has been renamed as eosinophilic granulomatosis with polyangiitis and is reviewed in the chapter on myocarditis (see Chap. 63).

• Toxic DCM, both exogenous and endogenous, with the former being mostly caused by drug toxicity leading to structural and persistent myocardial damage and dysfunction and the latter being represented by the presence of noxae that affect the heart. The proof of causality is essential to label the disease as toxic DCM.

• Inflammatory cardiomyopathies that are clinically defined by the presence of diagnostic criteria of DCM and pathologic evidence of myocardial inflammation (see Chap. 63).

• Peripartum DCM that manifests during the last month of pregnancy or in the 5 months postpartum (see Chap. 103).

The current nosology of the above entities to nongenetic DCM is subject to future modifications; in fact, new genetic diseases are being described in the setting of autoinflammatory Mendelian disorders and in PPCM. Rare “autoimmune diseases” are now recognized as genetic diseases. Accordingly, the collection of family data and genetic counseling should be incorporated in the clinical workup of all patients with DCM to generate precise data for future research and progression of knowledge to better characterize the mechanism(s) underlying a particular DCM.

Introduction

The clinical manifestations of autoimmune/immune-mediated cardiac involvement are heterogeneous and include ischemic heart disease when epicardial or intramural coronary arteries are involved, valve disease with possible dysfunction, conduction defects, arrhythmias, pericarditis, and myocardial damage, either showing chronic low-grade inflammation and fibrosis or isolated fibrosis (see also Chap. 100). Pulmonary hypertension (see Chap. 74) can complicate autoimmune diseases.

Cardiac involvement can either follow a proven diagnosis of autoimmune diseases or be the first manifestation in undiagnosed patients. In this latter case, the cardiology workup should extensively investigate the entire spectrum of possible traits associated with each of the suspected diseases: ocular (ie, uveitis), cutaneous (ie, both diffuse cutaneous and limited cutaneous systemic sclerosis), respiratory (ie, pulmonary hypertension and fibrosis), thyroid (common in many autoimmune diseases), renal (either renal failure or acute crises in certain diseases), gastrointestinal (ie, esophagus in systemic sclerosis), and nervous and skeletal systems. Routine cardiologic evaluation includes history and physical examination, ECG, and transthoracic echocardiogram; further imaging investigations are planned according to the diagnosis and individual needs.

Familial clustering may manifest with the presence of different autoimmune diseases in more than one family member (Fig. 58–6). The emerging autoinflammatory genetic diseases constitute a novel clinical need for differential diagnoses; accordingly, collection of family data may also be clinically and scientifically useful for sporadic diseases.

Autoimmune diseases that commonly involve the heart and include DCM-like phenotype/HF in the spectrum of their possible cardiac manifestations are systemic sclerosis, rheumatoid arthritis, systemic lupus erythematosus (SLE), polymyositis/dermatomyositis, and Sjögren syndrome, as well as overlapping syndromes (Table 58–3). Iatrogenic “cardiomyopathy” can result from common medications used for autoimmune diseases; the cardiac phenotypes are restrictive/dilated in case of hydroxychloroquine toxicity¹⁰⁷ and dilated in case of tumor necrosis factor (TNF)-α inhibitors (eg, etanercept, infliximab, adalimumab; see Chap. 61). Iatrogenic cardiomyopathy should be distinguished from the intrinsic risk of HF directly related to the different autoimmune diseases.

Systemic Sclerosis (Scleroderma)

Introduction

Systemic sclerosis (SS) is a clinically heterogeneous disorder of the connective tissue characterized by fibrosis of the skin and internal organs, vascular abnormalities, and presence of autoantibodies against various cellular antigens.¹⁰⁸ In limited cutaneous scleroderma, fibrosis mainly affects the face, arms, and hands; patients may demonstrate Raynaud phenomenon years before the skin manifestation and may develop pulmonary hypertension (see Chap. 74). In diffuse cutaneous scleroderma, fibrosis involves large cutaneous areas and internal organs, including the heart. Cardiologists contribute to the clinical workup, differential diagnosis, characterization of cardiac involvement, treatment of the heterogeneous clinical phenotypes, risk stratification, and decision making in emergency (eg, cardiac tamponade) or in chronic management of the disease. In the past decade, better awareness of the disease and early diagnoses have contributed to improved care.¹⁰⁹

Epidemiology

Currently available data indicate a prevalence of SS ranging from 50 to 300 cases per 1 million persons and an incidence ranging from 2.3 to 22.8 cases per 1 million persons per year.¹¹⁰ This wide range of prevalence variability reflects the diagnostic complexity of multifactorial syndromes with overlapping traits and manifestations and the different diagnostic criteria currently used to establish a precise diagnosis.¹¹¹ Classification and diagnostic criteria are a matter of continuous revision by scientific societies and experts.¹¹² Overlap syndromes, as well as silent, asymptomatic, early phases of the disease, impact the prevalence data. Female-to-male ratio ranges from 3:1 to 14:1; susceptibility to scleroderma demonstrates ethnic variations.¹¹³

Causes and Major Pathology Features

SS is a sporadic multifactorial disease in the majority of cases. Environmental factors and occupational risk factors have been reported in patients with scleroderma-like phenotypes.^108,114 Less commonly, SS clusters in families where members may show different autoimmune diseases.¹¹⁵ These data suggest the contribution of genetic factors that could contribute with a small dose effect of multiple genes whose products are involved in innate and adaptive immunity, autoinflammation, cell signaling, extracellular matrix architecture, DNA or RNA degradation, and apoptosis or autophagy diseases.^116,117 Replication of genome-wide association studies independently confirmed the candidacy of many susceptibility genetic variants, including selected human leukocyte antigen (HLA) class II molecules.¹¹⁸ The maternal-fetal microchimerism hypothesis may help explain the higher prevalence of SS in females: according to this hypothesis, fetal and maternal lymphocytes can cross the placenta during pregnancy and trigger a graft-versus-host-like microreaction that may culminate in scleroderma.¹¹⁹ This hypothesis is further supported by the presence of allogeneic cells in peripheral blood and in skin biopsy samples obtained from patients with scleroderma,¹²⁰ and by the short interval between pregnancy and onset of the disease, in particular in the first year after delivery.¹²¹

Interstitial fibrosis is the main pathologic feature observed in hearts affected by scleroderma (Fig. 58–7). Early features are microvascular damage and mononuclear cell infiltrates. In later stages, the main pathology features are dense fibrosis of the dermis, loss of interstitial cells and vasculature, and tissue atrophy.¹²² Table 58–4 summarizes pathologic tissue and cellular changes as well as the expression of possible markers in affected tissue and autoantibodies.

Cardiac Involvement

The heterogeneous clinical manifestations include myocarditis, pericarditis, and pericardial effusion; conduction disturbances; LV systolic and diastolic dysfunction; valve dysfunction; myocardial ischemia and coronary artery disease; and pulmonary hypertension.¹²³ Cardiac involvement (all manifestations) occurs in about 15% to 30% of patients with diffuse scleroderma.¹²⁴ A dilated cardiac phenotype (any cause) has been reported in up to 40% of patients with diffuse scleroderma.¹²⁵ However, because several potential causes may cause HF, the proportion of HF attributable to heart muscle involvement rather than to an alternative mechanism (eg, ischemic heart disease, valvular heart disease, or hypertension caused by scleroderma renal crises) is difficult to establish.¹²⁶ Myocardial fibrosis is common¹²⁷ and differs from that observed in ischemic heart disease; in fact, it does not occur in tributary areas of affected arteries, and hemosiderin deposits that are typically seen in post-ischemic myocardium are characteristically absent.¹²⁸ Low-grade inflammation can involve either small vessels or the myocardial interstitium. In patients with newly developed cardiac manifestations, myocardial inflammation can occur,¹²⁹ but with very low inflammatory burden. Involvement of cardiac vessels (capillaries, arteriolar and epicardial coronary arteries) can now be well assessed using CMR imaging and coronary angiotomography.¹³⁰ Pericardial involvement includes fibrinous or fibrous pericardial thickening, and focal adhesions and effusion are common; these abnormalities are often clinically silent and benign¹²⁸ (see. Chap. 66). In the majority of cases, small pericardial effusions do not cause clinical symptoms or impact prognosis.¹³¹ Hemodynamically significant pericardial effusions can be associated with HF; a small amount of rapidly accumulating pericardial fluid in the rigid, fibrosclerotic pericardium may occasionally cause tamponade.^131,132 The pericardial fluid is generally noninflammatory and does not contain autoantibodies, immune complexes, or evidence of complement depletion. Histologically, the pericardium shows fibrous thickening and nonspecific inflammation. Valve disease and pulmonary artery hypertension can be part of the complex cardiac involvement in scleroderma.^{123,126,128,132}

Conclusion

The increasing availability of disease- and organ-specific treatments targeting unique biologic networks and signaling pathways makes a precise diagnosis imperative: DCM-like end-phenotypes may benefit from common treatments for HF. Biologic therapies can target molecules involved in the mechanisms of the immune system, such as cytokines (TNF-α, interleukin [IL]-6), immune cells (B cells), or co-stimulation molecules (cytotoxic T-lymphocyte–associated antigen 4 [CTLA4]), and are currently used in several autoimmune diseases. These drugs provide an alternative to the existing treatment methods of disease-modifying antirheumatic drugs and other immunosuppressive medications.¹³³ Therefore, confounding scleroderma DCM with genetic DCM or other DCM phenotypes prevents or limits the identification of disease-specific causes.

Rheumatoid Arthritis

In patients diagnosed with rheumatoid arthritis (RA), DCM accounts for about 20% of mortality.¹³⁴ Compared to the general population, patients with RA still demonstrate higher mortality. In autopsy series, heart involvement was demonstrated in up to 60% of patients with RA.¹³⁴ As with other autoimmune diseases, RA hearts may show a variety of cardiovascular manifestations: DCM; pericarditis in up to 50% of cases¹³⁵ (but hemodynamically significant pericardial effusions is uncommon, 0.5% of patients); pericardial calcifications¹³⁶; nodular thickening and calcifications of cardiac valves¹³⁷; and coronary atherosclerosis.¹³⁴

Dilated Cardiomyopathy and Heart Failure

Congestive HF is a major cause of morbidity and mortality in RA patients.¹³⁸ The risk of HF remains high after adjustment for underlying coronary artery disease; in long-term follow-up studies, the burden of HF was significantly higher in RA patients than in a control population. Markers such as rheumatoid factor positivity and erythrocyte sedimentation rate, clinical traits such as extra-articular involvement, and treatments such as steroids are associated with risk of congestive HF after adjusting for coronary artery disease and risk factors, suggesting that RA is an independent risk factor for HF.¹³⁸ Diastolic dysfunction adds to systolic dysfunction in RA hearts.^139,140 The modification of cardiac mass in RA hearts is debated, with transthoracic echocardiography supporting this hypothesis versus CMR studies that did not confirm an increase in cardiac mass, but rather showed lower LV volumes compared with age-matched controls.¹⁴¹ The cause of diastolic dysfunction may be ascribed to the effects of inflammatory cytokines such as TNF-α, IL-1, and IL-6, all of which mediate inflammation and fibrosis; cytokine profiles seem to distinguish patients with moderate-to-severe diastolic dysfunction from those with normal heart function¹⁴²; higher myocardial levels of citrullinated proteins compared with controls may partly account for myocardial involvement in RA.¹⁴³

Iatrogenic Myocardial Damage Caused by Rheumatoid Arthritis Treatment

Disease-modifying drugs used to treat RA (and other connective tissue diseases) may lead to iatrogenic myocardial damage. Toxicity may result from long-term use of TNF inhibitors (TNF-I) and antimalarial agents such as hydroxychloroquine.^107,144,145 Hydroxychloroquine toxicity manifests with either hypertrophic-restrictive cardiomyopathy or DCM and is diagnosed with endomyocardial biopsy demonstrating myelin figures and curvilinear bodies (Fig. 58–8) (see Chaps. 61 and 100). The role of TNF-I in causing HF is debated. Etanercept and infliximab trials concordantly showed unfavorable outcomes.^146,147 However, other studies/registries reported that the prevalence of congestive HF is lower or comparable in treated patients versus nontreated patients; a meta-analysis of data from these registries concluded that the incidence of congestive HF is lower in treated than in nontreated patients.¹⁴⁸ In the most-recent guidelines (2015) for treatment of RA, the Voting Panel of the American College of Rheumatology reported that recommendations are conditional because the evidence is of very low quality and noted that “there are no reports of exacerbation of HF using non-TNF biologics and the US Food and Drug Administration (FDA) warns against using TNF-I in this population based on worsening of congestive HF with TNF-I in the Adverse Event Reporting System database. TNF-I should only be used if there are no other reasonable options, and then, perhaps, only in compensated HF”¹⁴⁹ (see Chap. 100).

Reactive Amyloidosis

Reactive amyloid A deposition occurs in RA as a result of long-term, uncontrolled inflammation. The rate of pathologic detection is 21% to 30% of patients with RA, with cardiac involvement in 28%.¹⁵⁰ Cardiac amyloidosis causes a restrictive phenotype with myocardial wall thickening and biventricular enlargement and atrial dilation. The diagnosis is suspected by imaging and by demonstration of involvement of other organs (renal or gastrointestinal tract, heart, spleen, liver, adrenal glands, and, less frequently, lungs, pancreas, thyroid, aorta, muscle, synovial membranes, lymph nodes, peripheral nerves, bones, and skin). The incidence of amyloidosis is likely related to the severity and duration of inflammation. Biologic therapy with anti-TNF agents may reverse amyloid deposition.¹⁵¹

Accelerated Atherosclerosis

Vascular heart disease may contribute to HF. Accelerated atherosclerosis is a major cause of morbidity and mortality in RA. A meta-analysis in 41,490 patients with RA from 14 studies showed a 48% increased risk of cardiovascular events compared with the general population.¹⁵² Both classic risk factors (diabetes, dyslipidemia, obesity, hypertension, and smoking) and disease-related factors, such as RA disease duration, RA positivity, and disease activity, contribute to increase the risk of cardiovascular diseases.^152,153 Paradoxically, decreased lipid levels may precede the diagnosis of RA,¹⁵⁴ but evolve during and after treatments, when cholesterol increases in particular in treated and responder patients.¹⁵⁵ TNF-I increase total cholesterol and high-density lipoprotein levels without an equivalent effect on low-density lipoprotein levels¹⁵⁶; statins may be required in treated patients with high low-density lipoprotein levels,¹⁵⁷ and discontinuation of statins in RA can increase the risk of myocardial infarction.¹⁵⁸

Rheumatoid Arthritis Heart Can Display a Pan-Cardiac Involvement

Cardiac involvement in RA can clinically manifest with cardiomyopathy, pericarditis, and valve disease, and can further be complicated by the development of reactive amyloidosis. Iatrogenic myocardial damage is proven for hydroxyl-chloroquine, but not confirmed for biologic drugs such as TNF-I, in particular regarding the development of HF. Several disease-specific risk factors and effects of medications add to traditional risk factors in the pathogenesis of the accelerated coronary atherosclerosis. Cardiovascular risk seems to decrease when combined methotrexate and TNF-I achieve a good control of the disease.¹⁵⁹

Systemic Lupus Erythematosus

Dilated Cardiomyopathy in Systemic Lupus Erythematosus

SLE is a multiorgan autoimmune disease demonstrating higher prevalence in African Americans than in whites, and typically affects women of childbearing age.^160,161 The clinical course is characterized by cyclic evolution, with periods of quiescence alternating with periods of disease activity. The heart is commonly involved: DCM is one of the most serious forms of organ involvement in SLE,^160,161 and LV systolic dysfunction is associated with poor outcome.¹⁶² In 2012, a consensus group of experts on SLE (the Systemic Lupus International Collaborating Clinics) revised the prior American College of Rheumatology criteria¹⁶³ for SLE and established the diagnostic criteria (Table 58–5).^164,165 Myocardial involvement is not included in diagnostic criteria; pericarditis is listed in the context of serositis and occurs in 10% to 30% of patients.^160,166 Although myocardial involvement does not provide a diagnostic contribution, it is common in early and late phases of the disease; diastolic dysfunction is detected in 45% of patients without evidence of cardiac disease¹⁶⁷; it also accounts for the majority of deaths.^168,169

Epidemiology

Most SLE manifests between the ages of 16 and 50 years; estimated incidence rates in North America, South America, and Europe range from 1 to 23 per 100,000 per year.^{169,170,171,172} Pediatric-onset SLE represents 10% to 20% of all SLE cases and is associated with greater disease severity than adult-onset SLE; the majority of pediatric-onset SLE patients will have developed irreversible disease manifestations within 5 to 10 years of disease onset, most commonly involving the musculoskeletal, ocular, renal, and neuropsychiatric systems.¹⁷⁰

Pathogenesis

The current pathogenic hypothesis integrates the effects of both environmental factors and genetic predisposition; risk factors are both SLE specific and traditional. Women are more commonly affected; the role of estrogens (17β-estradiol) and related receptors has been investigated. Estrogens are involved in the regulation of cellular subsets of the immune system through estrogen receptor–dependent and –independent mechanisms¹⁷³; estrogen receptors regulate innate immune cells and signaling pathways,¹⁷⁴ including all subsets of T cells and related cytokines,¹⁷⁵ and B cells, influencing B-cell differentiation, activity, function, and survival.¹⁷⁶

Epigenetic factors, specifically hypomethylation of CpG sites within genes involved in different pathways, seem to be associated with increased production of autoantibodies (anti-dsDNA, anti-SSA, anti-Sm, and anti-RNP antibodies)¹⁷⁷ and also influence endothelial-inflammatory cell interactions and inflammatory responses. Among susceptibility alleles identified with genome-wide association studies, NCF2, encoding a core component of the multiprotein NADPH oxidase, confers susceptibility to SLE in individuals of European ancestry¹⁷⁸; HLA-DRB1*15-DQB1*06 haplotype has recently been proposed as the greatest risk factor for development of disease in certain ethnic groups.¹⁷⁹ In a recent meta-analysis, the HLA-G 14-base pair insertion/deletion polymorphism was associated with susceptibility to a subgroup of autoimmune diseases such as SLE, but not RA.¹⁸⁰ Additional genes/loci have been assigned to SLE: PTPN22, FCGR2A, FCGR2B, CTLA4, TEX1, and DNAS1 are all included in the Mendelian Inheritance in Man (MIM) catalogue as associated with increased susceptibility to SLE. However, at present, there is no evidence supporting a Mendelian inheritance of SLE, which thus remains a potential multifactorial autoimmune disorder.

Myocardial Involvement

Heart and vessels are commonly involved in SLE. The major pathology features include accelerated atherosclerosis, myocardial fibrosis, and HF.

Accelerated atherosclerosis¹⁸¹ is not explained by traditional risk factors, but likely promoted by systemic inflammation.¹⁸² Traditional risk factors (eg, diabetes, hypertension, hyperlipidemia) are common in young SLE patients, often as a side effect of immunosuppressant therapy.¹⁸³ SLE-related risk factors include the presence of autoantibodies, such as antiphospholipid antibodies, anticardiolipin antibodies, and lupus anticoagulant,¹⁸³ which are produced in 30% to 40% of SLE patients. These antibodies do not fully explain the increased thrombotic risk in SLE; in fact, published data show that only 10% of “positive” patients experience a thrombotic event, and 40% of SLE thrombosis cases are in patients without such antibodies.¹⁸⁴ Treatment-related risk includes the effects of both corticosteroids and immunosuppressant drugs.¹⁸⁵ Risk stratification and modification for cardiovascular events should, therefore, be tailored on individual risk profile, based on modifiable risk factors and SLE-specific risk factors. Atherosclerosis is associated with increased risk of cardiac, cerebral, or peripheral arterial ischemic events and mortality; in a systematic review and meta-analysis including 17,187 patients from 32 studies, the incidence of cardiovascular events was almost 25%, and 4.5% of the patients had an acute myocardial infarction.¹⁸³ Although early subclinical atherosclerosis can manifest as intima-media thickening of carotid arteries, recent studies demonstrated that atherosclerotic plaque lesions can be found frequently in the absence of intima-media thickening in both SLE and SS patients; sonography of carotid and femoral arteries may identify additional atherosclerotic lesions and detect patients at a high risk for cardiovascular events.¹⁸⁶ Coronary contrast enhancement by CMR may detect subclinical disease in the coronary vessel wall, thus providing a novel direct marker of vessel wall disease.¹⁸⁷ Considering that traditional Framingham cardiovascular risk factors do not fully explain the excess of the risk of coronary artery disease in SLE,¹⁸¹ it is likely that inflammatory and autoimmune mechanisms interact with genetic, environmental, and treatment-related factors.

Myocardial fibrosis represents a significant independent predictor of adverse cardiac outcomes in both ischemic and nonischemic cardiomyopathies.^188,189 Recent CMR studies confirmed that mid-wall myocardial fibrosis is frequent in SLE and is associated with age, but not with disease duration or severity; in addition, late gadolinium enhancement comprising > 15% of LV mass may be associated with diastolic dysfunction and impaired exercise capacity.¹⁹⁰

DCM/HF is common in patients with SLE; although the causes can be related to the systemic (cytokine burden) and local inflammation, the role of accelerated atherosclerosis is not easy to dissect from local damage potentially induced by chronic inflammation. However, SLE patients with cardiomyopathy have a prognosis similar to those with idiopathic DCM.¹⁹¹

The European League Against Rheumatism recommends baseline cardiovascular evaluation and regular, yearly cardiovascular follow-up to assess smoking, vascular events (cerebral/cardiovascular), physical activity, oral contraceptives, hormonal therapies, and family history of cardiovascular disease; to perform blood tests, including blood cholesterol and glucose; and to measure blood pressure and body mass index (and/or waist circumference).¹⁹²

Definition

Precise diagnostic criteria of PPCM are essential to distinguish potentially reversible LV dilation and dysfunction from DCM with different causes, both genetic and nongenetic; pregnancy may in fact unmask an underlying genetic DCM (see also Chap. 103). In the workshop held by the National Heart, Lung, and Blood Institute and the Office of Rare Diseases (2000), PPCM was defined as a “rare life-threatening cardiomyopathy of unknown cause that occurs in the peripartum period in previously healthy women: diagnosis is confined to a narrow period and requires echocardiographic evidence of LV systolic dysfunction,”¹⁹³ According to the European Society of Cardiology (ESC), “PPCM is an idiopathic cardiomyopathy presenting with HF secondary to LV systolic dysfunction towards the end of pregnancy or in the months following delivery, where no other cause of HF is found. It is a diagnosis of exclusion.”¹⁹⁴ The LV may not be dilated, but the ejection fraction (EF) is nearly always reduced below 45%.¹⁹⁴ Both definitions share the following concepts: (1) idiopathic condition (no detectable cause of HF); (2) temporal appearance (the last month of pregnancy or during the first 5 months postpartum in the former and toward the end of pregnancy or in the months following delivery in the latter); (3) the absence of preexisting known heart disease; and (4) the presence of LV dysfunction. Therefore, the precise clinical diagnosis of PPCM should be based on these criteria. A further condition that could confirm the diagnosis is the regression of PPCM within 12 months after delivery.¹⁹⁴

Incidence

PPCM is estimated to occur in 0.1% of all pregnancies. The prevalence of PPCM ranges between 1:4000 and 1:1000 live births in the Western world.^195,196 Approximately 75% of cases are diagnosed within the first month after delivery, and 45% occur in the first week. Regional variation is wide: 1 case per 20,000 live births in Japan¹⁹⁷; 1 case per 1000 live births in South Africa¹⁹⁸; 1 case per 40,322 in China¹⁹⁹; and 1 case per 300 live births in Haiti.²⁰⁰ A high prevalence in Nigeria is likely caused by the tradition of ingesting kunun kanwa (dried lake salt) while lying on heated mud beds twice a day for 40 days postpartum to stimulate breast milk production.²⁰¹ This variable prevalence suggests that regional and cultural factors influence the risk of PPCM.

Risk Factors

Risk factors for PPCM include age (> 30 years), multiparity, African American race, obesity, substance use, pre-eclampsia, and chronic hypertension.^202,203 It is unclear whether race is an independent risk factor or whether the increased risk of PPCM results from the interaction of race with hypertension. Factors such as poor nutrition and lack of prenatal care are included in the list of risk factors, but are unconfirmed. Climate may also contribute: PPCM seems to be more common in tropical regions with increased heat and humidity.²⁰⁴

Etiologic Hypotheses

Etiologic hypotheses include familial/genetic predisposition, maladaptive response to hemodynamic stresses of pregnancy, viral myocarditis, stress-activated cytokines, malnutrition and selenium deficiency, fetal micro-chimerism, prolactin, and prolonged tocolysis²⁰⁵ (Fig. 58–9).

Genetic Hypothesis

In a recent study including 172 women (one-third of African descent) with PPCM versus 332 patients with DCM, sequencing of 43 DCM genes identified 26 heterozygous truncating variants in eight genes (17 mutations were identified in the TTN gene). The prevalence of truncation-predicting variants in PPCM (15%) was similar to that in DCM (17%) and significantly higher than in the reference population (4.7%); 65% of the truncating variants were in the TTN gene.²⁰⁶ This study may have relevant implications for both PPCM and genetic DCM; in fact, TTN mutations have been reported in 25% of FDCM and in 18% of nonfamilial, sporadic DCM.³² Persistence of the DCM phenotype in one-fifth of patients suggests the pathogenic role of the TTN mutation and implies that one-fifth of cases of PPCM are actually genetic DCM unmasked by pregnancy.³² Conversely, if the phenotype regresses after pregnancy among women with mutated PPCM, the role of TTN variants could be questioned not only in PPCM but also in DCM associated with the same variants. At present, it appears prudent then to consider a genetic predisposition to PPCM rather than a genetic cause.

Abnormal Hemodynamic Response

During pregnancy, in particular in the last trimester, blood volume and cardiac output increase and afterload decreases because of relaxation of vascular smooth muscle cells; LV mass and arterial compliance also increase, whereas total vascular resistance decreases.²⁰⁷ The consequent functional and transient LV “hypertrophy” aims at meeting fetal and maternal demands. In normal pregnancies, this transient LV functional and structural remodeling regresses shortly after birth.²⁰⁷ Although PPCM might, in part, be a result of an exaggerated decrease in LV function when these hemodynamic changes of pregnancy occur, this hypothesis is unlikely to explain, by itself, PPCM.

Viral Myocarditis and Systemic Inflammation

Past studies proposed viral myocarditis as the main mechanism for PPCM.²⁰⁸ This hypothesis remains unconfirmed. Clinical outcomes do not seem to differ in women with and without clinically diagnosed myocarditis. Increased levels of plasma inflammatory, apoptosis mediators and cytokines measured in women with PPCM do not correlate with LV function or outcomes. LVEF at the time of initial diagnosis remains the strongest predictor of outcome.²⁰⁹

Malnutrition and Selenium Deficiency

Deficiencies in selenium and other nutrients have also been advocated as causes of PPCM.²¹⁰ Selenium deficiency (serum selenium concentration < 70 μg/L) is a well-known cause of DCM and of increased susceptibility to viral infections, hypertension, and hypocalcemia.²¹¹ Selenium deficiency is a risk factor for PPCM in Kano, Nigeria, and seems to be related to rural residency.²¹¹

Fetal Micro-Chimerism

During pregnancy, the trafficking of circulating cells generates bidirectional micro-chimerism (ie, fetus to mother and vice versa). Cell trafficking from fetus to the mother starts very early; the numbers of fetal cells increase in the maternal circulation in pregnancies affected by pre-eclampsia and before the onset of pre-eclampsia symptoms.²¹² Adult women may bear chimeric blood cells from multiple sources.²¹³ For instance, microchimerism is one of the factors advocated in the pathogenesis of autoimmune diseases, which are more common in females than in males. The small shared epitopic-mimetic compound QKRAA has been proposed as a contributor to the pathogenesis of RA by acting as a ligand that activates pro-arthritogenic signal transduction events.²¹⁴ Women who lack RA susceptibility HLA alleles encoding the small QKRAA epitope can acquire RA susceptibility alleles from microchimerism, thus increasing their risk of RA.²¹⁵

Prolactin

Prolactin is one of the key candidates in the pathophysiology of PPCM. Increased levels of prolactin in pregnancy are associated with increased blood volume and levels of circulating erythropoietin and hematocrit and decreased blood pressure, angiotensin responsiveness, and levels of water, sodium, and potassium. In the last trimester of pregnancy, oxidative stress is enhanced; under defective antioxidative mechanisms, prolactin is increasingly cleaved into a 16-kDa fragment (16-kDa prolactin). This fragment may induce microvascular damage, leading to cardiac injury and dysfunction. The dopamine D₂ receptor agonist bromocriptine inhibits prolactin release and may represent a therapeutic option. This hypothesis was tested in a murine model that develops a PPCM-like phenotype; in this mouse model, the administration of bromocriptine prevented the development of the phenotype.^216,217 After successful testing in pilot series and single cases, a large clinical trial was planned and is ongoing, aimed at investigating the safety of the dopamine D₂ receptor agonist bromocriptine and its effects on LV function in women with PPCM.²¹⁸

Prolonged Tocolysis

Prolonged use of tocolytic therapies has been associated with development of pulmonary edema in pregnant women²¹⁹; based on this observation, chronic use of β-sympathomimetic medications was linked with PPCM. Prolonged tocolysis refers to the use of β-sympathomimetic drugs for more than 4 weeks. Although these agents are used for the management of various other conditions, the association between tocolytic therapies and HF appears to be unique to pregnancy.

Diagnostic Workup and Management

Echocardiography is the first-line imaging test. Ideally, baseline data should be available and demonstrate normal baseline LV size and function. Such baseline data, however, are rarely available because echocardiography is not indicated routinely in pregnancy and the later development of PPCM is unpredictable. However, early diagnosis is possible, and when treated in the early phases, LV systolic function may normalize in a high percentage of cases.¹⁹⁴

The increasing awareness of PPCM is contributing to earlier diagnosis and better management and outcomes. A major limitation is the absence of baseline echocardiographic evaluation in women who develop PPCM. However, the combined treatment with β-blockers plus angiotensin-converting enzyme (ACE) inhibitors or angiotensin receptor blockers is now recommended (Class I) by the AHA and ESC Guidelines^17,220 for treatment of HF with reduced LVEF after delivery. ACE inhibitors and angiotensin receptor blockers are highly teratogenic and should be avoided during pregnancy.²²¹ After delivery, PPCM should be treated in accordance with the current ESC guidelines for HF. During pregnancy, treatment options include hydralazine, long-acting nitrates, and β₁-selective β-blockers; diuretics (furosemide and hydrochlorothiazide) should be used sparingly because they can cause decreased placental blood flow¹⁹⁴ (see Chap. 103).

Recovery of Left Ventricular Function in Peripartum Cardiomyopathy

Although data reporting prognosis and outcome in PPCM differ (Table 58–6), most recent data concordantly demonstrate that LV function recovers in a significant percentage of cases (23% to 78%).^{205,222,223,224} In a recent prospective study in 100 patients from 30 US and Canadian centers (Investigations of Pregnancy-Associated Cardiomyopathy [IPAC]), full recovery (LVEF ≥ 50% at 12 months postpartum) occurred in 72% of patients. Outcomes were unsatisfactory for 13% of women, who had major events or evolved through severe chronic cardiomyopathy (LVEF < 35%) during the 1-year observation time after delivery. Women of African heritage had more frequent hypertension (preeclampsia, gestational, and chronic) and a later PPCM diagnosis.²²⁵ The most impressive data are from the German PPCM registry, with improvement in 85% of patients; the highest percentage of recovery (96%) was observed in patients treated with bromocriptine, β-blockers, and ACE inhibitors/angiotensin receptor blockers, albeit full recovery with an EF > 55% was present in only 47% of patients.²²⁶ Event-free survival was significantly worse for women with a baseline LVEF < 30% compared with those with LVEF > 30% (1-year event-free survival rate, 82% vs 99%; P < .004) (Fig. 58–10).²²⁶ Results recorded in the IPAC registry in women who did not receive bromocriptine were comparable with those reported by German investigators in PPCM patients who received bromocriptine in addition to standard evidence-based treatment.²²⁶

Introduction

Light to moderate alcohol intake is beneficial for cardiovascular health,²²⁷ whereas habitual heavy alcohol consumption is associated with increased risk of LV dilatation and dysfunction (alcoholic cardiomyopathy [A-CMP]),²²⁸ arrhythmias,^228,229 systemic hypertension,²³⁰ ischemic heart disease,²³¹ stroke,²³² and skeletal muscle abnormalities.²³³ Heavy alcohol consumption corresponds to daily ingestion of at least three standard-size drinks or alcohol consumption over 80 g; light-moderate alcohol intake corresponds to less than three standard-sized drinks or < 80 g/d. Because standard-sized portions of wine, liquor, or beer contain approximately the same amount of alcohol, the daily amount of ethanol is usually measured per number of standard drinks.^228,229

Alcoholic Cardiomyopathy

A-CMP clinically manifests with DCM phenotype; it is considered a specific disease by both the ESC and AHA.^8,9 In the ESC consensus document on the classification of cardiomyopathies, A-CMP is included among the acquired forms of DCM.⁹ Thiamine deficiency (beriberi) is no longer confused with A-CMP; the former fully responds to thiamine administration, whereas the latter does not.²²⁸

Among alcoholics, the prevalence of DCM is 0.43% in women and 0.25% in men.²³⁴ An excessive ethanol intake is reported in 3% to 40% of patients with DCM.²³⁵ The three criteria for diagnosing A-CMP include DCM phenotype, absence of other known and detectable causes of DCM, and a long history of heavy alcohol intake. The toxic effect is expected for daily alcohol consumption over 80 g (three or more standard-sized drinks per day) lasting 5 years or more before the onset or diagnosis.^234,235,236 The symptoms correspond to those of DCM, with patients presenting with HF as a result of reduced cardiac output and atrial and/or ventricular arrhythmias. The LV function ameliorates with either abstinence or reduction of the daily intake of alcohol.^{228,234,235,236} There are no specific guidelines for the treatment of A-CMP; current management should include complete alcohol abstinence and therapies recommended to treat HF as a result of other causes (see Chap. 70).

Mechanisms

The toxic damage of ethanol is attributed to a nonoxidative metabolic pathway for alcohol related to fatty acid metabolism in the heart, skeletal muscle, pancreas, and brain.^228,229 Factors that have been involved in the pathogenic mechanisms of A-CMP include apoptosis, abnormal excitation-contraction coupling, functional and structural mitochondrial damage, loss of contractile filaments, deregulation of protein synthesis, ATPase, L-type calcium channels, and activation of the renin-angiotensin and sympathetic nervous systems.^228,229 None of these factors and mechanisms has been confirmed, and they may all contribute. However, alcohol-induced mitochondria toxicity ends in mitochondrial dysfunction, loss of mitochondrial membrane potential and increase in mitochondrial oxidative stress, eventually culminating in cell death.²³⁷ Nonselective mitochondrial damage could link oxidative stress, apoptosis, and mitochondrial loss, and explain the multiorgan damage in those with alcohol use disorders.²³⁷ Recent experimental studies showed that chronic plus binge ethanol feeding in mice causes alcohol-induced cardiomyopathies characterized by increased myocardial oxidative/nitrative stress, impaired mitochondrial function and biogenesis, and enhanced cardiac steatosis.²³⁸

Prognosis

Current knowledge regarding A-CMP is mostly based on historical studies; data on prognosis using contemporary HF pharmacologic agents are not available. Alcohol withdrawal is associated with A-CMP regression in a 40% to 50% of cases, but not in all of them, suggesting that factors other than alcohol abuse may contribute to the natural history of the disease.²²⁸ A-CMP seems to have a better prognosis than idiopathic DCM; during a median follow-up of 59 months (interquartile range, 25-107 months), 94 consecutive patients with A-CMP demonstrated higher transplantation-free survival than 188 patients with nonalcoholic DCM. During follow-up, 63% of patients reported remaining abstinent, 32% continued alcohol consumption but had reduced intake to < 80 g/d, and only 5% were persistent or heavy alcohol drinkers (> 80 g/d). Atrial fibrillation, QRS width > 120 ms, and the absence of β-blocker therapy identify patients with a poor outcome. ACM patients who reduced their alcohol intake to moderate levels exhibited similar survival (P = .22) and cardiac function recovery (P = .8) as abstainers.²³⁹

Definition and Impact of the Diagnosis

Patients with cirrhosis frequently demonstrate hyperdynamic circulation and increased cardiac output, decreased systemic vascular resistance, and increased compliance of the arterial vessels. This combination of hemodynamic conditions corresponds to latent left HF where any stressor may clinically unmask the LV dysfunction.^240,241 According to the 2005 World Congress of Gastroenterology, cirrhotic cardiomyopathy is defined by “chronic cardiac dysfunction in patients with cirrhosis characterized by impaired contractile responsiveness to stress and/or altered diastolic relaxation with electrophysiological abnormalities in the absence of other known cardiac disease.”²⁴¹ The impaired diastolic relaxation is more common in early phases, whereas systolic dysfunction and LV dilation are manifest in advanced phases; electrophysiologic abnormalities include prolongation of QT interval in more than 50% of cases, electromechanical dyssynchrony, and chronotropic and/or inotropic incompetence^240,242 (Table 58–7). Cardiac evaluation is increasingly relevant for patients who are candidates for liver transplantation and/or transjugular intrahepatic portosystemic shunt (TIPS) implantation. These may exacerbate the hyperkinetic circulation and compromise cardiac function. TIPS decompression of the portal hypertension increases the preload to the heart, pulmonary artery pressure, and systemic vascular resistances, thus lowering afterload. In the complex pathophysiology of cirrhosis-related cardiac and hemodynamic changes, diastolic dysfunction seems to predict death after TIPS implantation.²⁴³

In addition, pretransplant diastolic dysfunction predisposes patients with hepatic cirrhosis to post-transplant adverse cardiac events.²⁴⁴ Of 107 patients who received liver transplantation, 24% developed HF in association with diastolic dysfunction prior to liver transplantation, and diastolic dysfunction was a predictor of a poor prognosis.²⁴⁵ Cardiac causes of immediate death after liver transplantation include postreperfusion syndrome, pulmonary hypertension, and cardiomyopathy.²⁴⁶ Therefore, cirrhotic cardiomyopathy may influence the evolution of post-TIPS and orthotopic liver transplantation; however, it does constitutes a contraindication to both procedures, in particular to liver transplantation.²⁴⁷

Pathogenesis

Hepatic failure and portal hypertension have been considered as possible factors for the development of cardiac changes in patients with cirrhosis. In the early phase of cirrhotic disease, the hemodynamic changes are compensated by the development of a hyperdynamic circulation. In later phases, as splanchnic arterial vasodilation progresses, effective arterial blood volume declines, and there is subsequent activation of the renin-angiotensin-aldosterone system and development of LV dilation and dysfunction. Factors involved in the development of cardiac impairment include inflammation mediated by intestinal bacterial translocation,²⁴⁸ cytokines that can influence both LV remodeling and function, and the negative inotropic effects of nitric oxide, carbon monoxide, and endogenous cannabinoids.²⁴⁹

Diagnosis and Imaging

Cardiac evaluation of patients with cirrhosis requires special attention to clinically latent signs of HF, as it may influence candidacy for TIPS or liver transplantation. The presence of diastolic dysfunction should be recognized before liver transplant because it may be associated with intraoperative complications and may complicate the early postoperative course.²⁵⁰ Latent cardiac dysfunction, defined by abnormal stroke volume response to unclamping of portal vein, is common in liver transplant recipients.²⁵¹ Preoperative low systemic vascular resistance and left atrial dilation are independent predictors of inadequate responses.²⁵¹ All cardiac imaging modalities contributing to better characterize diastolic dysfunction should be integrated to provide best assessment of diastolic dysfunction; echocardiography with tissue Doppler is the most common method to detect and monitor diastolic dysfunction. A few studies have applied CMR and contrast-enhanced CMR as well as cardiac CT with coronary artery calcification scoring. More studies are needed to establish the clinical contribution of both CMR and CT in patients with cirrhotic cardiomyopathy.²⁵² Tests with either physiologically or pharmacologically induced circulatory stress are useful to assess systolic dysfunction; echocardiography with tissue Doppler is the most common method to detect diastolic dysfunction (E/A and E/E′ ratio). Strain techniques can offer incremental information. The impaired cardiac pharmacologic response to stress can be additionally investigated with magnetic resonance myocardial stress testing; deformation parameters may be more sensitive in identifying abnormalities in inotropic response to stress than are conventional methods.²⁵³ Cirrhotic cardiomyopathy (CCR) is, therefore, assuming an increasing relevance considering the negative impact that an unrecognized CCR may have during and after liver transplantation. An early diagnosis of CCR with pretransplant treatment of HF may prevent an acute onset or worsening of cardiac failure after liver transplantation.²⁵⁴ Pretransplant treatment of HF may improve quality of life before transplantation and reduce perioperative complications after liver transplantation.

Different types of DCM that have been variably included in past classifications of cardiomyopathies are potentially reversible. In future nosologies of cardiomyopathies, these entities should probably be separated from cardiomyopathies and be part of a disease-specific classification. Apart from PPCM, A-CMP and CCR, other DCMs that potentially reverse when the cause is controlled and treated are tachycardia (PVC)-induced cardiomyopathy, premature ventricular contraction–induced cardiomyopathy, takotsubo cardiomyopathy, metabolic cardiomyopathies (both endocrine and nutritional), cardiomyopathies complicating end-stage chronic diseases such as renal failure (uremic cardiomyopathy), and inflammatory and infectious cardiomyopathy. These latter are reviewed in Chap. 63. The existence of an independent “obesity cardiomyopathy” characterized by systolic dysfunction fulfilling criterion for the diagnosis of DCM is debated; recent studies highlight the role of obesity as an independent risk factor for the development of HF, but obesity-related HF is thought to be multifactorial.²⁵⁵

Tachycardia-Induced Cardiomyopathy

Tachycardia-induced cardiomyopathy is caused by persistent tachycardia that induces increased ventricular filling pressures, severe biventricular systolic dysfunction, reduced cardiac output, and increased systemic vascular resistance. Arrhythmias include atrial fibrillation, atrial flutter, atrial tachycardia, reentrant supraventricular tachycardia, accessory pathway tachycardia, frequent ectopic beats, and ventricular tachycardia; sustained rates above 115 or 120 bpm may be an important prognostic factor.²⁵⁶ Sinus tachycardia has also been reported as a cause of cardiomyopathy.²⁵⁷ The diagnosis is based on the absence of other causes of nonischemic cardiomyopathy (eg, hypertension, alcohol or drug use, stress) and on the absence of LV hypertrophy and relatively normal LV diameter (LV end-diastolic dimension < 5.5 cm). The recovery of LV function after control of tachycardia (by rate control, cardioversion, or radiofrequency ablation) within 1 to 6 months confirms the diagnosis.^258,259

Premature Ventricular Contraction–Induced Cardiomyopathy

PVC–induced cardiomyopathy is characterized by LV dilation and/or dysfunction, diastolic dysfunction, and ventricular strain by speckle tracking imaging.²⁶⁰ A dose-response relation has been demonstrated in serial evaluations of LV function among 239 consecutive patients with frequent PVCs and no obvious cardiac disease; > 20,000 PVCs per 24 hours were associated with subclinical deterioration in LVEF, whereas > 10,000 PVCs per 24 hours showed LV dilation without a change in LVEF.²⁶¹ An increasing PVC burden was associated with adverse outcomes in 1139 healthy participants of the Cardiovascular Health study (≥ 65 years) with an initial normal LVEF and no history of congestive HF who underwent 24-hour Holter monitoring. Over a median follow-up of > 13 years, a PVC burden in the upper quartile was associated with a threefold greater odds of a decrease in LVEF, a 48% increased risk of incident congestive HF, and a 31% increased risk of death, compared with the lower quartile.²⁶² PVC-induced cardiomyopathy appears reversible with reduction in ectopy burden. Rapid normalization of LV parameters follows successful radiofrequency ablation of the PVC focus.^263,264 After successful ablation, the rates of improvement in LVEF have varied between 47.1% and 100%.^263,264 Randomized trials are ongoing, aimed at assessing whether therapy for frequent PVCs in patients with idiopathic LV dysfunction modifies clinical outcomes. The Early Elimination of Premature Ventricular Contractions in Heart Failure study (ClinicalTrials.gov identifier: NCT01757067) is testing the effects of ablation versus medical therapy; the Reversal of Cardiomyopathy by Suppression of Frequent Premature Ventricular Complexes study (NCT01566344) is testing conventional HF therapy and PVC suppression therapy (ablation or amiodarone if unsuccessful ablation) versus HF therapy alone. In children, the proportion of PVC-induced cardiomyopathy seems higher than previously expected especially because ectopy tends to persist throughout follow-up. The possibility of radiofrequency ablation of the ectopic ventricular focus could be considered in children who manifest LV ventricular dilation or dysfunction.²⁶⁵ The pathophysiology of PVC-induced cardiomyopathy remains unknown, and the optimal management of these patients is not yet established.

The Left Apical Ballooning Syndrome

The left apical ballooning syndrome, or takotsubo cardiomyopathy, is a transient hypocontractility of the mid and apical segments of the left ventricle associated with hyperkinesis of the basal walls. This mismatched contractility causes a balloon-like appearance of the distal ventricular walls in systole. The phenomenon may mimic acute coronary syndrome in the absence of coronary artery disease or spasm (see Chap. 36). The transient dysfunction is most prevalent in older women exposed to emotional or physical stressors. The disease can manifest with atypical phenotypes such as reversed and right ventricular takotsubo and global hypokinesis.^266,267 Biomarkers include increased levels of circulating catecholamines, especially epinephrine. Takotsubo cardiomyopathy may not follow obvious physical or emotional stress; in addition, it may carry significant early and late serious complication risks. It can occur concomitantly in the presence of subcritical coronary artery disease; patients often demonstrate cardiovascular risk factors and associated comorbidities.²⁶⁷ Primary takotsubo cardiomyopathy is frequently related to emotional stress, is largely reversible, and carries a benign prognosis; in contrast, secondary forms are associated with acute potentially severe conditions and may result in a higher event rate.²⁶⁸ Although included in the cardiomyopathy classification, they do not share the chronic morphologic LV abnormalities and remodeling that characterize DCM.

Metabolic Cardiomyopathies

Metabolic cardiomyopathies are secondary to effects of primary noncardiac defects of energy production leading to impaired cardiac function. They may be caused by several different endocrine diseases and nutritional deprivations. Endocrine cardiomyopathies include growth hormone (GH) excess and GH deficiency and thyroid hypo- and hyperfunction.^{269,270,271,272,273,274}

Excess or Deficiency in Growth Hormone and/or Insulin-Like Growth Factor 1

Excess or deficiency in GH and/or insulin-like growth factor-1 (IGF-1) is deleterious for the cardiovascular system. Acromegaly is an endocrine disease with specific somatic changes caused by an excess of GH; in the majority of cases, the cause is a pituitary tumor producing GH. Chronic GH and IGF-1 excess causes a cardiomyopathy that is characterized by early concentric cardiac hypertrophy associated with diastolic dysfunction and by later evolution to systolic dysfunction.²⁶⁹ The control of GH/IGF-1 excess is accompanied by a decrease in cardiac mass and improvement of cardiac function.²⁷⁰ In patients with hypopituitarism, either childhood onset or adulthood onset, GH deficiency is characterized by decrease of cardiac mass and impairment of systolic function; other effects include increased cardiovascular risk with changes in body composition, unfavorable lipid profile, insulin resistance, endothelial dysfunction, and atherosclerosis. In patients with GH deficiency, the administration of recombinant GH is followed by increase in cardiac mass, improvement in cardiac performance, and improvement of the cardiovascular risk factor profile.^270,271 Therefore, in acromegaly and GH deficiency, the control of GH/IGF-1 secretion reverses cardiomyopathy and risk factors and restores normal life expectancy.²⁷¹

Thyroid Cardiomyopathy

Thyroid dysfunction may be associated with either high cardiac output cardiomyopathy in hyperthyroidism or low cardiac output cardiomyopathy in hypothyroidism. Thyroid hormone effects are exerted on β-adrenergic receptors, contractile apparatus, ion-ATPase pump, and the sarcoplasmic reticulum.²⁷² Clinically, thyroid function is usually investigated in patients that manifest high-output HF. Hyperthyroidism is characterized by resting tachycardia and increased blood volume, stroke volume, myocardial contractility, and EF. High-output HF may be a result of a tachycardia-mediated cardiomyopathy mechanism.²⁷³ In hypothyroidism, thyroid hormone deficiency is associated with by bradycardia, decreased myocardial contractility and relaxation, and prolonged systolic and early diastolic times. Cardiac preload is decreased as a result of impaired diastolic function, cardiac afterload is increased, and chronotropic and inotropic functions are reduced. Subclinical hypothyroidism is associated with increasing risk of cardiovascular events, but not of total mortality.²⁷⁴ Medical or surgical treatment of either hyper- or hypothyroidism restores cardiac function.²⁷⁴ Accordingly, thyroid cardiomyopathy is potentially reversible. Recurrent pericarditis and pericardial effusion are well-known complications of hypothyroidism (see Chap. 66); less commonly, pericarditis is observed in thyrotoxic crisis in Graves disease.²⁷⁵ Two medical emergencies, myxedema coma and thyroid storm, are unrelated to thyroid cardiomyopathy, but may involve cardiologists for the diagnosis and management. Myxedema coma is characterized by somnolence, lethargy, and hypothermia. It is precipitated by triggers such as cold exposure, infection, drugs trauma, stroke, HF, and gastrointestinal bleeding, and is more common in old women presenting with altered consciousness and history of hypothyroidism, neck surgery, or radioactive iodine treatment (www.thyroidmanager.org). Thyroid storm presenting with palpitation, atrial arrhythmias, and hypertension is managed according to a multidisciplinary approach aimed at lowering thyroid hormone synthesis and secretion and circulating thyroid hormones and controlling the peripheral effects of thyroid hormone up to the resolution of systemic manifestations.²⁷³

Uremic Cardiomyopathy

Uremic cardiomyopathy is a common complication in chronic renal diseases and is considered a risk factor for morbidity and death among patients with chronic kidney disease (CKD) or end-stage renal disease.²⁷⁶ Along with coronary artery disease, LV hypertrophy and congestive HF account for approximately 50% of deaths in patients with CKD. Nontraditional risk factors (eg, uremic toxins, renin-angiotensin-aldosterone system activation, sympathetic nervous system activation, anemia, calcium phosphate imbalance, inflammation) are emerging as factors involved in the pathogenesis of cardiac disease in CKD. Their specific role in uremic cardiomyopathy is not clarified.^277,278 A recent Kidney Disease Improving Global Outcomes report emphasized the need for unraveling the origin and mechanism of cardiac dysfunction and diagnosing asymptomatic LV dysfunction in patients with CKD.²⁷⁹ Compared with conventional hemodialysis (3 days per week), frequent hemodialysis (nocturnal home hemodialysis, short daily hemodialysis, and peritoneal dialysis) optimizes the ultrafiltration rate with minimal reductions in intravascular volume and cooling the dialysate; this reduces dialysis-induced myocardial stunning and intradialysis hypotension. Kidney transplantation may improve or reverse uremic cardiomyopathy and confers a significant survival advantage over hemodialysis.²⁸⁰ Early diagnosis of uremic cardiomyopathy contributes to risk stratification and decisions about proper type of dialysis therapy.²⁸¹

Introduction

Cardiovascular toxicity during and after chemotherapy and radiotherapy for cancer has been and remains a potential limitation to leveraging the full benefits of most advanced treatments of malignancies in childhood and adulthood (also see Chap. 101).²⁸² Progression in cancer treatments has significantly increased the number of long-term survivors, but has also increased the possibility of toxicity manifesting to the cardiovascular system over time.^283,284 Cardiovascular toxicity includes a range of complications, including involvement of the myocardium (eg, DCM, myocarditis), pericardium (eg, pericarditis, pericardial effusion), and vascular system (eg, coronary atherosclerosis, peripheral arterial disease, acute aortic syndromes, coagulopathy, hypertension, and hypotension)^285,286 (Table 58–8). Although these possible complications are known, the individual risk of occurrence remains largely unpredictable. Whereas greater risk would be expected in older patients and those with a preponderance of traditional risk factors, clinical evidence across a range of ages indicates that underlying risk does not necessarily correlate with expression of phenotype. Pretreatment risk stratification requires meticulous consideration to prevent or limit the effects of potentially harmful treatments in patients at high risk while not excluding patients from receiving effective treatments.²⁸⁷ New cardiovascular imaging tools (ie, strain imaging) and biomarkers (eg, troponin, B-type natriuretic peptide) are now contributing to early detection, better risk stratification, and monitoring of the cardiovascular toxic effects of chemotherapy.^288,289 None of these markers has been specifically generated for detection or monitoring of cardiovascular toxicity per se; rather, their use represents the extension of standards in cardiovascular diagnosis generally to chemotherapy-induced cardiovascular abnormalities specifically.

The relevance of oncocardiotoxicity and DCM is related to the paradigmatic and well-known anthracycline-induced DCM, and to the possibility that patients with genetic DCM may develop cancer. A detailed discussion of cardiovascular toxicity induced by chemotherapy is provided in Chap. 101.

Myocardial Toxicity

Patients who develop cancer may have either healthy cardiovascular systems or preexisting cardiovascular conditions, including DCM (see Table 58–8). The former is obviously the most common, favorable, and expected condition. The latter requires involvement of a cardiologist, multidisciplinary evaluation, and specific monitoring.^{284,286,287,290,291,292} Heart muscle toxicity is generally characterized as nonreversible injury (type 1) as a result of the presence of structural damage (prototyped by the anthracycline or high-dose cyclophosphamide DCM in acute and chronic forms) and potentially reversible (on cessation of therapy) dysfunction (type 2) in the absence of structural abnormalities, as with targeted therapies (ie, trastuzumab, sunitinib, lapatinib).^{291,292,293,294,295,296,297} In patients first treated with drugs causing type 1 injury and then with drugs causing type 2 injury, the damage may be cumulative. Myocardial toxicity occurs with many agents used in the treatment of cancer; examples are summarized in Table 58–9.^298–334

Cancer Therapy-Related Cardiac Dysfunction

The concept of myocardial toxicity does not coincide with that of cardiomyopathy.¹ The substantial difference between the diagnostic criteria in toxic myocardial dysfunction related to cancer therapy (American College of Cardiology Foundation/AHA 5.4.3) versus nontoxic dysfunction (eg, inherited DCM) is the dynamic concept of the diagnosis. The cardio-oncotoxicity is a double-step diagnosis that implies normal baseline cardiovascular function and evidence of decline of function during or after treatment. Therefore, the definition of cancer therapy–related cardiac dysfunction (CTRCD) or chemotherapy-related cardiac dysfunction (CRCD) seems to better describe both type 1 and type 2 myocardial damage. Anthracyclines (see Table 58–9) are the best-known drugs increasing the risk of CTRCD caused by type 1 injury; the incidence rate is as high as 2.1% in randomized controlled trials and 26% in observational studies.²⁸² Targeted drugs such as trastuzumab, a monoclonal antibody targeting the human epidermal growth factor receptor-2 in breast or gastroesophageal malignancies, are associated with CTRCD caused by type 2 injury with 3-year cumulative incidence of 6.6% when trastuzumab is used with anthracyclines and 5.1% when used without anthracyclines. Higher risk in mid- and long-term follow-up is reported with trastuzumab after anthracycline treatment.^282,284,285 The possibility of CTRCD occurring in FDCM may confound the interpretation of the cause of the myocardial disease (Fig. 58–11).

In 2002, the Review and Evaluation Committee of Trastuzumab-Associated Cardiomyopathy supervising trastuzumab trials defined CRCD as follows: “(1) characterized by a decrease in cardiac LVEF, either global or more severe in the septum; (2) symptoms of HF; (3) associated signs of HF, including but not limited to S3 gallop, tachycardia, or both; and (4) decline in LVEF of at least 5% to less than 55% with accompanying signs or symptoms of HF, or a decline in LVEF of at least 10% to below 55% without accompanying signs or symptoms. The presence of any one of the four criteria is sufficient to confirm a diagnosis of CRCD.”³¹⁷ From 2002 to the present, the above criteria have been repeatedly discussed and applied in clinical trials, but their systematic application is still debated and guidelines for their application are still lacking.²⁸²

The type 1 or 2 classification of myocardial injury is not immediately evident, and the diagnosis of type 2 injury can only be made after cardiac abnormalities have resolved following cessation of chemotherapy. The definition of type 1 or 2 injury cannot be made a priori. Reversibility or nonreversibility of the CRCD requires evidence of regression or persistence of myocardial damage (as with imaging, pathology, biomarkers, or clinical evaluation).

Myocarditis

Myocarditis is a rare complication described in patients treated for cancer (ie, cyclophosphamide and hemorrhagic myocarditis). It may present as fulminant myocarditis.^335,336

Pericardial Involvement

Pericardial effusion and pericarditis (excluding carcinomatous pericarditis) may occur commonly in patients treated with different classes of chemotherapeutic drugs. They may be reliable markers of cardiotoxicity, especially when it can be demonstrated that the pericardial abnormality was not present prior to the administration of chemotherapy (see Chaps. 66 and 101).

Rhythm Disturbances

Variations of cardiac rhythm and ECG intervals are common. Prolongation of the QT interval is a consequence of many classes of anticancer drugs²⁸⁶ and, according to Common Terminology Criteria for Adverse Events (CTCAE) version 4.03, is graded from 1 to 4 (the grade 5 corresponds to death) based on the length of the corrected QT interval (http://evs.nci.nih.gov/ftp1/CTCAE/CTCAE_4.03_2010-06-14_QuickReference_5x7.pdf). The risk of malignant arrhythmias in patients who develop a prolonged QT interval is not necessarily the same as in patients with congenital prolonged QT syndromes. There are no studies comparing populations of patients with long QT syndromes and toxic prolongation of the QT interval. Nonetheless, drugs that prolong the QT interval should be discontinued when the QT interval approximates warning values indicated by the guidelines. Arrhythmias, both supraventricular and ventricular, can be subtle and may escape detection because they can be clinically silent, transient, and unperceived by patients. Once life-threatening arrhythmias are demonstrated in patients treated for cancer, exclusion of noncardiotoxic causes is essential to prevent an incorrect assignment of the event to the chemotherapeutics, to decide about premature termination of chemotherapy, and to appropriately manage the arrhythmias.

Arterial Blood Pressure

Chemotherapeutics may either increase or decrease blood pressure (see Table 58–9). Hypertension is common in adult patients; it should be either excluded or recognized with certainty before assigning the hypertensive complication to chemotherapy.³¹⁵ The CTCAE version 4.03 grading of hypertension includes four grades (http://evs.nci.nih.gov/ftp1/CTCAE/CTCAE_4.03_2010-06-14_QuickReference_5x7.pdf); grade 1, or prehypertension (systolic blood pressure of 120-139 mm Hg or diastolic blood pressure of 80-90 mm Hg), is very common in the general adult population (see Chaps. 23 and 24). In nonhypertensive patients, chemotherapy-induced endothelial dysfunction is associated with loss of vasorelaxant effects and suppression of anti-inflammatory and vascular repair functions. These effects may initiate and further maintain the development of hypertension, thrombosis, and atherogenesis. Hypertension is the most common cardiovascular complication associated with vascular endothelial growth factor inhibitors.³³⁷ Effective antihypertensive treatment is mandatory before starting chemotherapy. In addition, drugs such as ACE inhibitors may exert a nonspecific protection to cardiotoxicity (see Chap. 25).^283,286,289

Vascular Toxicity

Atherosclerosis-related events such as ischemic heart disease (both acute and chronic) and coagulation imbalances and related events (both thrombosis and hemorrhage) can be difficult to manage if a patient’s vascular profile was unknown prior to initiation of treatment. Current protocols do not include definition of coronary anatomy or evaluation for coronary artery disease before prescribing chemotherapy in asymptomatic patients. A pretreatment risk assessment with either noninvasive evaluation of coronary tree (eg, angiographic CT coronary evaluation) in high-risk patients or evaluation of surrogate markers of atherosclerosis (combining risk factors with ultrasound evaluation of carotid arteries) could help cardiologists identify patients with different classes of risk, avoiding excess of vascular monitoring while selecting high-risk patients for more strict follow-up. Dynamic vascular reactions such as coronary spasm or Raynaud phenomenon are also described as an expression of vascular toxicity of several chemotherapeutics. Finally, rare cases of aortic dissection have been reported in patients treated with sunitinib,^338,339 gemcitabine,³⁴⁰ sorafenib,³⁴¹ and bevacizumab.³⁴²

Overall, cancer therapy related cardiovascular toxicity may affect myocardium function and electrical properties of the heart, pericardium, aorta, and peripheral vessels and may also include coagulation imbalances. The role of cardiologists is increasingly complex as new classes of drugs enter clinical cancer programs and complications expand from CTRCD to electrical complications with variable risk of life-threatening arrhythmias. Cardio-oncology units should regularly operate in all cancer centers and units. (The list of drugs and current protocols is available and regularly updated at http://www.cancer.gov/about-cancer/treatment/drugs/.)

The prognosis of DCM has been largely investigated in nongenotyped series of nonischemic DCM index patients. Prognostic stratification largely coincides with that of systolic HF (see Chap. 70). Currently available data are based on the phenotype and do not systematically include the etiology. Exceptions include small series of genetic dilated cardiolaminopathies and dilated cardiodystrophinopathies, potentially reversible DCM, PPCM, and DCM in autoimmune diseases.

In the overall population of DCM patients, the prognosis has progressively and significantly improved over the past three decades.^343,344,345 The optimized use of ACE inhibitors, angiotensin receptor blockers, β-blockers, antialdosterone therapy, and resynchronization therapy^17,291 has substantially modified the natural history of DCM (see Chaps. 68 and 70). Annual incidence of major cardiac events (death and cardiac transplantation) declined, and transplant-free survival improved. HF-related deaths have decreased by 40% in the past 20 years.³⁴⁶ Reasons for better survival and low event rate include early diagnosis, earlier medical treatment, resynchronization therapy, and primary prevention of sudden death with ICD implantation^347,348 (see Chaps. 72 and 73).

Recent studies demonstrate that patients enrolled in recent decades had a shorter history of HF, were less symptomatic for HF, and had fewer previous hospital admissions for HF and less severe heart disease; this modification has partly been attributed to the beneficial impact of systematic familial screening.³⁴⁷ The enrollment in clinical series of family members (identified as affected in family screening studies) added a proportion of patients who were diagnosed in the early phases of the disease. In a single-center series, the incidence of major events declined to < 2 per 100 patients per year, with an 87% survival free from heart transplantation at 8 years in the last decade.³⁴⁷ Earlier diagnoses imply earlier medical treatments, which are consequent further contributors to improved clinical outcome.

The pharmacologic and device-based actions modifying LV remodeling contribute to improvement of LVEF. Key drugs such as ACE inhibitors and β-blockers^17,291 and the use of aldosterone antagonists, improving LV remodeling and systolic function, positively impact prognosis^349,350,351; up to 50% of patients receiving cardiac resynchronization therapy demonstrate significant reverse remodeling during the course of 24 months in nonischemic DCM with more evident benefits when follow-up is longer and patients demonstrate mild HF.³⁵² In patients in New York Heart Association class I/II, cardiac resynchronization therapy decreased all-cause mortality, reduced HF hospitalizations, and improved LVEF.³⁵³

Early reverse remodeling, achieved with both medical and device-based therapy, positively influences decisions on primary prevention, preventing unnecessary ICD implantation in patients with idiopathic DCM presenting with SCD in Heart Failure Trial criteria.³⁵⁴ Reverse remodeling further influences mitral valve dysfunction, which is a further relevant negative prognostic contributor.³⁵⁵ MitraClip can improve symptoms and promote reverse remodeling in nonresponders to cardiac resynchronization therapy.³⁵⁶ The proportion of super-responders to cardiac resynchronization therapy in DCM series is difficult to establish; however, the predictors of responsiveness to cardiac resynchronization therapy can be usefully applied to DCM, especially because they include the lack of prior myocardial infarction.³⁵⁷

Overall, current prognostic markers and optimized medical and interventional treatments provide evidence that the natural history of DCM substantially improved in the last decades and can be further ameliorated. The open problem remains the arrhythmogenic risk stratification. Even super-responders to cardiac resynchronization therapy remain at risk for ventricular arrhythmias.³⁵⁸ Imaging can contribute to predict arrhythmogenic risk; the association of myocardial fibrosis with mortality and SCD in patients with DCM is known.^359,360 However, beyond fibrosis, prognostic predictors such as low LVEF have limited sensibility and specificity. Selecting patients according to the current guidelines shows that most DCM patients do not actually benefit from ICD implantation, while suffering collateral effects. In addition, it is now evident that patients exhibiting mildly depressed LVEF can be at risk of sudden death. In the complex substrates underlying SCD, multiple risk factors can probably contribute to predict the risk of SCD.⁶² Genetics can significantly contribute to stratify arrhythmogenic risk, as demonstrated in LMNA mutation carriers^59,60 and by the first introduction of genetics in the 2015 ESC guidelines for the management of patients with ventricular arrhythmias and prevention of SCD.⁶³ Overall, we expect that in time the increased proportion of genotyped patients with DCM will generate data useful for progressing from the phenotype-based prognostic stratification to cause-specific risk stratification, either genetic or nongenetic.

The current descriptive diagnosis of DCM is based on a common end-phenotype that is functionally characterized by systolic dysfunction and LV remodeling with either mild or severe LV dilation. During the course of the past five decades, several different disorders, of myriad origins and causes, have been categorized as DCM simply based on shared phenotype. This “nosology strategy” was mainly based on the paucity of information about specific causes and pathophysiologic mechanisms of the different types of diseases that could culminate in a DCM phenotype and on the available, overlapping guideline-based treatments that influence the phenotype independent of the underlying mechanism. However, knowledge gleaned with novel DNA sequencing technologies, biomarkers, and advanced imaging is demonstrating that different causes of DCM call for different diagnostic and treatment strategies. The majority of DCMs have heterogeneous genetic origins, with more than 100 disease genes discovered to date. Other diseases with end-DCM phenotype are phenocopies of genetic DCM; each subtype is opening novel avenues of disease-specific research and paving the way for novel, targeted treatments. Typical examples are uremic cardiomyopathy and CCR. DCM secondary to metabolic diseases or other potentially treatable causes, such as persistent tachycardia, may reverse with the control or treatment of the primary cause. It is time to separate DCM by cause, generating novel nosology and descriptors useful to improve disease-specific and individually tailored monitoring and treatments. Confounding different etiologies in the broad group of phenotypes is an obstacle to the development and delivery of personalized treatment. Cardiomyopathies in general and DCM in particular can generate precise disease models exportable to other cardiovascular diseases.