Cardiomyology: an attempt to link structural cardiac and skeletal muscle damage in patients with dilated cardiomyopathy

Abstract

Background

Patients with chronic heart failure syndrome may develop a myopathy contributing to muscle wasting and exercise intolerance. Little is known about skeletal muscle pathology in patients with dilated cardiomyopathy (DCM).

Aims

To correlate skeletal muscle biopsy (SMB) findings with endomyocardial biopsy (EMB) and clinical/functional data in DCM patients.

Methods and results

SMBs and EMBs were morphometrically and morphologically analysed in 30 consecutive patients aged 40.7±14 years (16–63), including 25 in NYHA classes I and II. Four had familial DCM. Serum creatine-phosphokinase (sCPK) was normal in 23 and slightly increased in seven.

All the SMBs showed morphological (including non-recurrent changes different from those in the EMBs) and morphometric changes, with atrophic fibres in all cases (atrophy factor >150 in six NYHA II patients) and necrotic fibres in three. The SMBs of two patients with EMB-proven myocarditis showed inflammation. Plotting the morphometric factors against age at symptom onset, age at diagnosis, follow-up, clinical outcome, ejection fraction and sCPK levels showed that only sCPK >180 mU/ml correlated with atrophy factor.

Conclusions

The use of SMB in DCM patients documents some aspecific subclinical muscle damage that is unrelated to the functional class and duration of the DCM. However, this information does not contribute to identify the aetiology or managing the disease.

Keywords Skeletal muscle biopsy; Dilated cardiomyopathy; Endomyocardial biopsy

Introduction

Patients with chronic heart failure (CHF) syndrome may develop a related myopathy that contributes to muscle wasting and exercise intolerance.^1,2 A small subgroup of CHF patients diagnosed as having idiopathic dilated cardiomyopathy (DCM) seems to have sub-clinical skeletal muscle involvement, only testified by a persistently increased serum creatine-phosphokinase (sCPK) levels,^3–5 which are considered a hallmark of neuromuscular disease^6,7 but may also be associated with physical exercise,⁸ and endocrinal,⁹ toxic, infectious and neoplastic conditions.¹⁰

DCM is not synonymous of CHF. Little is known about the pathology of skeletal muscle in patients with DCM, with the exception of five DCM and four hypertrophic cardiomyopathy cases showing myogenic changes,¹¹ X-linked DCM,^12,13 rare mitochondrial diseases,^14,15 rare familial cardiomyopathies associated with clinically overt myopathies,^16–18 and myositis in patients with enterovirus-related myocarditis.¹⁹ On the other hand, the pathology of skeletal muscle in CHF patients (any origin) has been extensively investigated,^20,21 and the functional profile of CHF-related myopathy has been clarified.

Information concerning skeletal muscle in DCM patients could be clinically helpful if the disease-causing insult is shared by both types of striatal muscles. Most DCM patients do not show the clinical markers of myopathy, but informative data could be obtained from pathological studies of skeletal muscle biopsies (SMBs), an invasive but riskless procedure that can be performed surgically or by means of a muscle biopsy needle.²³ The likelihood that a SMB can contribute to diagnosis greatly depends on the clinical setting: in the myology field, a recent study of 114 candidates for myopathy on the basis of persistently increased sCPK levels and muscle weakness found that SMB allowed a definite diagnosis of myopathy in 10.5% of cases and a probable diagnosis in 7.9%.²² In DCM, the possible role of SMB is still far from being justified in terms of clinical benefit, but it can be speculated that pathological studies of myocardial and peripheral skeletal muscles may be of use.

This prospective study was designed to assess the presence of myopathic changes in consecutive DCM patients, to investigate the histomorphological, histoenzymatical and immunohistochemical characteristics of skeletal and cardiac muscle, and to correlate the changes found in skeletal muscle with those observed in the heart and with functional data. The final aim was to establish whether SMB findings may add information to the etiopathology of DCM.

Methods

A consecutive series of 30 patients diagnosed as having "idiopathic" DCM (Table 1) underwent SMB on the day preceding right heart catheterisation and endomyocardial biopsy (EMB). Patients with known neuromuscular disorders, cachexia, or diagnosed or clinically suspected infiltrative or specific heart disorders, as well as patients assuming statins were excluded.

The study was part of a project dedicated to DCM that was approved by the Ethics Committee of I.R.C.C.S. Policlinico San Matteo, Pavia, Italy. All of the patients were given detailed information concerning the aims and potential informativeness of the procedure: only those who showed that they had understood the aims and who gave their informed consent were included.

Diagnosis of dilated cardiomyopathy

DCM was diagnosed on the basis of the WHO criteria.²⁴ Coronary artery disease was excluded by means of coronary angiography. Familial forms were identified by screening all of the living, informed and consenting relatives by means of a clinical evaluation, electrocardiogram, echocardiogram and the measurement of sCPK levels.³

Endomyocardial and skeletal muscle biopsy

The EMBs were taken from the right ventricular apex via the internal jugular vein,^25,26 and the needle SMBs from the middle part of the vastus lateralis muscle. There were no procedure-related complications in either case. The SMB specimens were divided into three parts: two were cooled for 15 s in a isopentane bath immersed in −160 °C liquid nitrogen and kept at −80 °C for light microscopy histoenzymatic examination; the third was immersed in Karnowsky solution for electron microscopy. The histochemical and histoenzymatic reactions were performed on consecutively cut 10-mm thick frozen sections: the former consisted of hematoxylin-eosin, modified Gomori trichrome, PAS, PAS-diastase digestion and Oli Red O; the latter included: cytochrome c oxidase, succinate dehydrogenase, NADH dehydrogenase, adenosine triphosphatase (pH 4.2, 4.6, 10.6), alkaline and acid phosphatases, myoadenylate deaminase, non-specific esterase, myophosphorylase and phosphofructokinase. The immunohistochemical studies of both the endomyocardial and skeletal muscle biopsies included immunostaining with antibodies/antisera to T-lymphocytes (CD3, CD45R0), B-lymphocytes (CD20 and CD79a), macrophages (CD68), antigen-presenting cells and activated T-lymphocytes (HLA-DR), endothelial cells (CD31 and CD34), dystrophin (CO, NH, rod domain and mid-rod domain), α, β, γ, δ sarcoglycans, β-dystroglycans, merosin and spectrin, emerin and lamin A/C.^12,27 Calculations were made of the percentage distribution of type 1 and 2 fibres, diameter variability, and atrophy and hypertrophy factors in both fibre types. The mean diameter was considered reduced when it was less than the lower normal value.²⁸ The EMB and SMB ultrastructural studies were performed on 1–2 mm samples fixed in Karnovsky solution and routinely processed as previously described.¹⁹ The SMB morphological and morphometric evaluation was separately and blindly performed by two authors (E.A. and G.A.). Active, resolving and borderline myocarditis was diagnosed according to the Dallas criteria.²⁹

Molecular studies

The molecular studies were aimed at investigating dystrophin in all of the male patients; mitochondrial DNA in the patients with organelle ultrastructural abnormalities and/or biopsy-revealed histoenzymatic alterations in the mitochondrial complex; actin and desmin gene analysis in the four familial forms; and lamin A/C gene in the four patients with atrioventricular block in patients with atrial fibrillation and in patients with increased sCPK.^{12,27,30–33}

Statistical analysis

The continuous data are given as mean values±standard deviation (SD), and the categorical data as percentages. The continuous variables were compared using Student's t test for unpaired data at one-way analysis of variance with Bonferroni's correction, and the categorical variables by means of chi-square analysis. A p value of <0.05 was considered significant.

Results

Clinical characteristics

The study population consisted of eight females and 22 males aged 16–63 years (mean age: 40.7±14 years). Their individual clinical characteristics are shown in Table 1. The mean disease duration from symptom onset to biopsy was 2.2±3.6 years (range: 1–192 months). The duration of follow-up from the time of the biopsy to the last clinical control was 2.2±1.2 years (range: 1–49 months).

None of the patients were smokers or had lung disease at the time of diagnosis. Eight patients with a very recent disease onset were clinically suspected as having myocarditis on the basis of previous influenza-like episodes (n=5), and broncho-pulmonary illness (n=3); all potentially infective illness had resolved at the time of the onset of cardiac symptoms. The symptoms had appeared less than six months before the study in 16 patients and more than six months before in 14. Four of the patients were in New York Heart Association (NYHA) class I, 21 in NYHA class II, two in NYHA class III and three in NYHA class IV (all with a very recent onset, and two with clinically suspected myocarditis).

All but four of the patients (#8, #9, #21 and #24, Table 1) had been in stable clinical condition for at least one month before the study, and all were receiving standard medical therapy consisting of ACE-inhibitors (80%), vasodilators (4%), diuretics (55%), anticoagulants (55%; INR 2–3) or antiplatelet agents (11%), amiodarone (27%), digitalis (48%) or β-blockers (53%). None of the patients was taking statins.

Mean peak oxygen consumption (VO2) was 14.37 ml/kg/min (between class B and C of Weber scale).³⁴ The level of sCPK at rest was measured twice in all patients: it was normal (<180 mU/ml) in 23 patients, minimally increased in three (218, 248 and 257 mU/ml) and moderately increased in four (about three to five times the normal value: 688, 944, 697 and 500 mU/ml) (Table 1). Serum CK-MB levels were normal in all but the two patients with acute myocarditis (maximum acute phase value: 63 mU/ml).

Twelve of the 30 patients had left and one right bundle branch block; two patients with proven sporadic disease had first-degree atrioventricular block plus bundle branch block, and two (one with suspected myocarditis, NYHA class IV, and the other with NYHA class I sporadic DCM plus paroxysmal atrial fibrillation) had first-degree atrioventricular block without bundle branch block.

Pathology of the muscle and heart

The morphometric and morphological data were obtained by means of light microscopy, with further morphological data being obtained by means of electron microscopy.

Skeletal muscle: light microscopy study

The morphometric study included the measurement of atrophy and hypertrophy factors (Tables 2 and 3). All of the SMBs contained atrophic fibres, but the age/gender and muscle type normalised values showed that only seven patients (six in NYHA class II and one in NYHA class I) had a higher than expected atrophy factor (>150). The atrophy factor in the remaining 23 patients was <150 in both type 1 and 2 fibres. The outcome data are shown in Table 1.

Hypertrophic type 1 and 2 fibres were found in 10 biopsies. Three patients had necrotic fibres and sCPK values of, respectively, 257, 105 and 697 mU/ml: the first had acute enterovirus-related myocarditis, the second familial autosomal dominant disease of unknown origin, and the third (who had been originally been suspected as having myocarditis because of the very recent – three weeks – disease onset after a flu-like episode) was unexpectedly found to have a dystrophin gene defect. Centrally located nuclei (<3%) were occasionally seen in a further six cases.

Myocardial versus skeletal muscle tissue

1. Patients with and without increase sCPK

The EMBs of the patients with normal sCPK levels showed the aspecific features commonly seen in DCM: interstitial fibrosis, irregularly sized and shaped myocytes and loss of contractile myofibrils. Their corresponding SMB did not show fibre necrosis. Of the seven patients with sCPK levels of >180 mU/ml, three had borderline values (218, 257 and 248 mU/ml) and four had values of >500 mU/ml. The corresponding muscle showed necrosis in one of the latter, whereas the other two patients with single fibre necrosis had normal sCPK levels.

2. Patients with and without myocarditis

Conventional histopathology of the EMBs showed active lymphocyte myocarditis in two of the eight clinically suspected cases. The corresponding muscle biopsies showed endomysial inflammatory infiltrates consisting of HLA-DR+ T-lymphocytes (CD45RO) and macrophages (CD68), and non-specific changes at both light and electron microscopy. The EMB characteristics of a further patient with clinically suspected myocarditis (with normal sCPK) were consistent with repairing/healing myocarditis: sparse inflammatory infiltrates in fibrosis without any active myocyte damage. The corresponding skeletal muscle biopsy showed rare perivascular inflammation consisting of HLA-DR+ CD45R0 lymphocytes and CD68 macrophages. The remaining five patients with clinically suspected myocarditis showed no myocardial inflammation, and all five SMBs showed rare CD45RO and CD68 cells. Among the 22 cases without any clinical signs of myocarditis, no significant inflammation was documented in either the EMBs or SMBs.

3. Patients with histoenzymatic changes (Tables 2 and 3)

The histoenzymatic studies testing COX, SDH and NADH mitochondrial reactions showed focal COX defects in five biopsies (mean number of fibres: 1.7), occasional moth-eaten fibres in 12, granularity in two, and target-like patterns in one. There were no significant abnormalities in phosphorylase, phosphofructokinase, myoadenylatedeaminase, acid phosphatase, alkaline phosphatase or non-specific esterase, with the exception of three patients with necrotic fibres.

In one (#19) of the five patients with occasional COX-defects, ultrastructural studies revealed mitochondrial proliferation in both heart and skeletal muscle, with an abnormal size and cristae being observed particularly in the myocardium. Three patients were found to carry heteroplasmic mithochondrial DNA mutations (G3337A Val-Ile ND1, C1703T combined with A15902G tRNA^Thr, A5510C Leu-Phe ND2) with higher amount of mutated DNA in heart than in muscle and blood.

Ultrastructural study of heart and skeletal muscle (Tables 2 and 3)

The ultrastructural studies of the SMBs and EMBs documented subcellular abnormalities in all 30 cases, including sarcomeric fibril loss and lipid droplets (Fig. 1).

The differences between the ultrastructural features of the two types of striatal muscles in the same patient were as follows. The six patients with rod bodies in the heart had no rods in skeletal muscle (Fig. 2). The five patients with hyaline bodies in muscle did not have them in the heart. Of the twelve patients with subsarcolemmal mitochondria accumulation in the skeletal muscle, five showed mitochondrial changes and proliferation in the heart (Fig. 3). A subsarcolemmal localisation of mitochondria was typical of skeletal muscle, whereas in the cardiomyocytes proliferating mitochondria spreaded throughout the cells. The patient with a dystrophin defect showed severe cell membrane damage in myocardial and typical δ-lesions in skeletal muscle. The ultrastructural myocardial and skeletal muscle changes in the same patient were different regardless of whether the defect was genetic (dystrophin and mitochondrial DNA defects), infectious (the two cases of enterovirus-related myocarditis) or unknown (the remaining 23 patients). None of the patients showed defective immunostaining of the dystrophin-associated glycoproteins, emerin and lamin A/C. In particular, dystrophin immunostaining was not significantly changed in the two patients with enterovirus-related myocarditis. None of the 14 patients in whom the LMNA gene was analysed (#5, #6, #7, #9, #10, #12, #13, #17, #18, #20, #21, #22, #23, #26) had pathological mutations.

Familial diseases

There were five familial cases: one had autosomal dominant disease, two were sib-pairs, one had X-linked DCM (with the mother carrying the dystrophin defect), and one had a possible matrilinear inheritance.

This last case was one of the five patients with ultrastructurally abnormal mitochondria in the heart and muscle, and rare COX-negative myocardial fibres; she was found to carry the heteroplasmic A15902G (tRNA^Thr) and C1703T mutations of mitochondrial DNA. Her muscle biopsy also showed some variations in fibre size, rare nuclear internalisation, peripheral purple staining and occasional pre-ragged fibres upon modified Gomori trichromic staining, the absence of COX activity in a few fibres, and the absence of inflammation. The other two patients with mitochondrial DNA mutations were the only clinically affected members of their families. Their EMBs showed aspecific histopathological features, and their SMBs variations in muscle fibre size, one with occasional nuclear internalisation.

In the X-linked DCM caused by a dystrophin defect (deletion of exons 48–51), the carrier mother was healthy and the sister was not a carrier (if the defect in the proband had not been identified, this disease would have been diagnosed as sporadic). Direct DNA sequencing showed that none of the other four familial cases had actin or desmin gene defects. None of the patients with sporadic DCM plus first-degree atrioventricular block as well as none of these with increased sCPK and atrial fibrillation had lamin A/C gene defects.

Statistical analysis

The atrophy and hypertrophy factors were plotted against age at symptom onset, age at diagnosis, duration of follow-up, clinical outcome, ejection fraction and sCPK levels. Only a sCPK level of >180 mU/ml correlated with the atrophy factor (p=0.009 for type 1 and p=0.01 for type 2 fibres), but it did not identify cases with an atrophy factor of >150 (p=NS).

Discussion

The results of this study show that: (1) all DCM patients show subclinical muscle changes, the presence of which is unrelated to aetiology, disease duration or functional class; (2) the morphological changes caused by identical insults are different in peripheral and cardiac striatal muscles, as documented by dystrophin and mitochondrial DNA defects, and enteroviral myocarditis; (3) the diagnostic contribution of SMB was less than expected, even in patients with increased sCPK levels; and (4) the heart and muscle tissue morphometry/morphology data do not correlate with clinical status or other patient data.

DCM patients show subclinical muscle changes

The presence of pathological changes in all of the skeletal muscle biopsies supports previous rare field observations.¹¹ Regardless of NYHA functional class or aetiology, all DCM patients have muscle abnormalities. As most of the patients in our series were in a low functional class, it is unlikely that observed skeletal muscle changes were due to tissue damage related to the severity of CHF: as previously suggested by Caforio et al.,¹¹ the tropism of the insults damaging the myocardium probably extends to skeletal muscle but, for still unknown reasons, the clinical phenotype only manifests at cardiac level. This is true for patients with familial disease and known genetic defects (mitochondrial DNA mutations and one dystrophin gene defect), or myotrophic infections (the two enteroviral-related myocarditis with myositis), as well as for patients in whom the cause of the disease is unknown. In the study by Caforio et al.,¹¹ five patients with DCM underwent SMB but not EMB: all showed pathological muscle changes, but none of these changes contributed information concerning aetiology or the defect underlying the disease.

The presence of hyaline bodies in the skeletal muscle of five of our patients raises the hypothesis of a diagnosis recently described by Selcen et al. in two brothers with cardio-neuromyopathy.¹⁸ However, skeletal muscle hyaline bodies in our patients were not accompanied by rod bodies, which were found in five non-corresponding hearts. In nemaline myopathy associated with familial hypertrophic cardiomyopathy¹⁸ caused by sarcomere gene defects, both heart and muscle may be affected by a genetic defect whose major clinical phenotype manifests at cardiac level but also causes subclinical structural changes in the corresponding muscle.

None of the pathological features was recurrent in patients with high sCPK levels, and none could be considered a specific marker of muscle damage increasing sCPK (with the only exception of the case with the dystrophin defect).

Morphological changes in the two types of striatal muscle

A striking set of findings is the morphological differences between heart and muscle observed in the same patient: identical defect/insults (infections, dystrophin and mitochondrial DNA defects) led to different pathomorphological features in the two striatal muscles, especially at ultrastructural level. One of our major hopes in designing the study had been that of finding identical morphological markers in heart and muscle that would aid the identification of disease aetiology. A part from non specific changes such as myofibrillar loss and lipid droplets, both the light microscopy patterns and the subcellular changes revealed by electron microscopy were different in the two muscle types and did not provide useful aetiological insight. Identical insults seem to have different structural effects on the two types of striatal muscles and we therefore hypothesise that, in a restless heart, subcellular changes are influenced by its continuous activity: for example, proliferating mitochondria spread throughout cardiomyocytes but mainly concentrated in the subsarcolemmal areas of skeletal muscle. Furthermore, the δ lesions of necrotic muscle fibres in patients with dystrophin defects are not seen in cardiomyocytes, in which sarcolemmal damage is characterised by a loss of basement membrane and cell membrane fragmentation. The artefactual contraction bands that occur in EMB samples may limit the interpretation of the necrotic patterns related to membrane damage, but δ lesions are also absent in hearts excised at transplantation, in which no artefacts are produced by the microtome. If it is true that a restless heart has an impact on morphological changes in the two types of striatal muscle, we still need to learn which morphological changes are shared by the two types of muscles insulted by an identical trigger.

Diagnostic contribution of skeletal muscle biopsy

The SMBs did not provide any substantial diagnostic benefits in relation to those offered by EMBs and peripheral blood DNA analyses. However, the diagnostic contributions of muscle and heart biopsies has to be discussed considering the clinical setting. All of patients came to our attention because of CHF, and had any clinical signs of myopathy. The 17-year-old man with a dystrophin defect was well until two weeks before being referred to our hospital because of suspected myocarditis, and an EMB was specifically requested on the grounds of this clinical hypothesis. Furthermore, patient #6, who was suspected as having myopathy on the basis of his high sCPK levels and SMB results, did not show any corresponding clinical signs: he improved with medical therapy and is stable in NYHA class II 27 months after biopsy. Both cases are typical examples of how a misleading clinical presentation can affect clinical diagnosis (and probably clinical decisions) when no tissue or molecular studies are available. However, the SMBs did not make a diagnostic contribution in these or any of the other cases: the diagnosis of the dystrophin defect relied on peripheral blood DNA gene analysis and EMB immunostaining, the diagnoses of myocarditis were made on the basis of EMB inflammation, and the four patients with mitochondrial DNA defects were suspected as having mitochondrial DCM on the basis of EMB findings. The main reason for the lack of of SMB to make a diagnostic contribution is that, although the muscle samples of the remaining 23 patients showed abnormalities, most of the morphological changes were non specific. Accordingly, our SMB programme was stopped before the planned time because the results did not contribute to improving patient management. Although SMB is a riskless procedure (at least in comparison with EMB), we had to conclude that, as things stand, it is not appropriate even in the case of DCM patients with high sCPK levels because, in the cardiological setting, all of the data that can be obtained from skeletal muscle are better addressed by EMB for diagnostic purposes. A few selected cases of cardiomuscular disorder may benefit from SMB, but this indication must be evaluated by expert cardiologists and myologists on an individual basis.

Whether the structural abnormalities observed in NYHA class I and II patients regardless of the onset of symptoms represent early subclinical changes caused by CHF or are related to the aetiology of the disease (in patients whose lifestyles are minimally modified by the disease) does not seem to have any immediate clinical relevance.

Correlations

The only positive correlation in our study was between the atrophy factor of type 1 skeletal muscle fibres and increased sCPK levels. Although all of the skeletal muscle biopsies contained some atrophic fibres, they were more frequent in the patients with high sCPK levels. This finding suggests that, in addition to the possibility that higher sCPK levels are caused by necrotic fibres, muscle fibre atrophy may be associated with sCPK release. However, analysis of the individual cases showed that the increase in sCPK is not predicted by necrosis and vice versa ; and the same is true for atrophic fibres. The statistical results therefore only provide a trend that fits with the expected result, but does not predict the extent of muscle damage in individual patients.

Conclusions

The use of SMB in DCM patients documents some aspecific subclinical muscle damage that is unrelated to the functional class and duration of the DCM. However, this information does not contribute to identify the aetiology or managing the disease. SMB is therefore inappropriate in unselected consecutive patients with DCM.

References