Relationship of Cardiac Magnetic Resonance Imaging and Myocardial Biopsy in the Evaluation of Nonischemic Cardiomyopathy

Authors


Carl V. Leier, MD, Division of Cardiovascular Medicine, 473 West 12th Avenue, The Ohio State University, Columbus, OH 43210
E-mail: carl.leier@osumc.edu

Abstract

©2012 Wiley Periodicals, inc.

This study was performed to determine the relative role of cardiac magnetic resonance (CMR) imaging and endomyocardial biopsy (EMB) in the evaluation of cardiomyopathy. Sixty-six patients with a clinical diagnosis of nonischemic dilated cardiomyopathy or restrictive cardiomyopathy underwent both EMB and CMR imaging as part of their diagnostic evaluation. The authors retrospectively reviewed the results of these two methods to determine their diagnostic impact and congruency. CMR imaging provided data on cardiac anatomy, left ventricular volumes, mass, and function in 85% of the patients, uncovered fibrosis in 31%, myocardial ischemia in 7%, and fibrofatty infiltration in two patients. EMB provided the histologic findings of cardiomyocyte hypertrophy in 77% of patients and substantial interstitial fibrosis in 59%. Six patients had EMB-proven amyloid heart disease, which was detected by CMR imaging in two. CMR imaging showed patterns of late gadolinium enhancement supportive of infiltrative disease or inflammation in 6 patients with EMB-proven definite (n=3) or borderline (n=3) myocarditis, but failed to do so in two other patients with borderline and two with resolving myocarditis. At the present time, CMR imaging and EMB remain complementary procedures in the evaluation of cardiomyopathic conditions.

The utility of cardiac magnetic resonance (CMR) imaging in the evaluation of myocardial disease and heart failure is becoming increasingly well established at centers with appropriate facilities and expertise. CMR imaging with late gadolinium enhancement imaging (LGE) has become a standard technique in the assessment of cardiac anatomy and function along with providing clinically useful information regarding myocardial perfusion, viability, ischemia, infarction, and fibrosis.1–7 CMR imaging offers diagnostic and prognostic information when applied to certain forms of dilated and infiltrative cardiomyopathy.1,2,4,5

Endomyocardial biopsy (EMB) has traditionally served as the method to establish the diagnosis of myocarditis and infiltrative cardiomyopathies, although only attaining a class IIa indication as a diagnostic technique in most recent heart failure guidelines.8 Its general use is also limited by sampling error9,10 and the risks associated with an invasive procedure including perforation of right heart structures.

It is, therefore, reasonable to pursue diagnostic methods that are less invasive yet provide similar or better diagnostic information at an acceptable level of sensitivity. Preliminary studies and reports indicate that CMR imaging is such a method,1,11–13 although uncertainty remains regarding its utility compared with EMB.1 The current study sought to determine the relative role and merit of CMR imaging and EMB and their level of congruency in the evaluation of nonischemic cardiomyopathy.

Methods

Study Population

All patients who underwent both EMB and CMR imaging for the evaluation of what was clinically deemed to be nonischemic cardiomyopathy after June 1, 2004, were included in this retrospective study. The 66 patients consisted of 39 men and 27 women with an age range of 18 to 81 years. Information regarding their demographic profile, basic laboratory data, coronary angiography, and the pre-CMR and pre-EMB clinical impressions is provided in Table I. Pertinent laboratory studies, when available, included biomarkers, troponin level, sedimentation rate, and C-reactive protein value (Table I).

Table I.   Demographic, Laboratory, and Angiographic Information and Clinical Indications for Study
Patient No.Age, ySexLaboratoryCoronary ArteriographyClinical Indication for EMBClinical Indication for CMR
Troponin IESRCRP
  1. Abbreviations: ARVC, arrhythmogenic right ventricular cardiomyopathy; CABG, coronary artery bypass graft surgery; CM, cardiomyopathy; CMR, cardiac magnetic resonance imaging; CHF, congestive heart failure; CRP, high-sensitivity C-reactive protein; CTA, computed tomographic angiography; EMB, endomyocardial biopsy; ESR, erythrocyte sedimentation rate; NA, not available; NP, not performed; PCI, percutaneous coronary intervention; VT, venous thrombosis.

123Male361512NPRestrictive CMCM, infiltrative
235Male<0.15NANANormalDilated CMCM
356Female<0.1585NANPRestrictive CMRestrictive CM
439Female1.4947100NormalRestrictive vs pericardialPericardial disease
534Male0.321NADiffuse plaquesDilated CMCM
629FemaleNANANANPDilated CMCM
747Male<0.15NANAIsolated plaqueDilated CM; VTVentricular arrhythmia
818MaleNA33NPRestrictive CM vs pericardialChest pain, pericardial disease
959MaleNANANANPRestrictive CMCM
1080Male0.2NANANPRestrictive CMChest pain, infiltrative process
1122Male<0.15NANANormalRestrictive CM; pulmonary sarcoidosisInfiltrative process
1229Male<0.1511Normal by CTADilated CMAngina
1341Female12.66546NormalDilated CMCM
1442Male<0.15NANANormalDilated CMCM
1531Male5.3117499NormalDilated CMCM
1653Female<0.1571NormalDilated CMCM
1759Female<0.15112NANPDilated CMCoronary atherosclerosis
1875Female<0.15NANANormalDilated CMCM
1967Male<0.1553NAIsolated plaqueRestrictive CM vs pericardialPericardial disease
2046Male0.4696185Isolated lesion (50%)Dilated CM; VTVentricular arrhythmia
2148Male<0.15NA48NPRestrictive CMInfiltrative CM
2234Male25.115Isolated plaqueDilated CMCM
2379Female<0.15NANANPDilated CMCM
2449Male<0.1584Isolated plaqueARVC; VTVentricular arrhythmia
2530Female1.173012NormalDilated CMCM
2637Male<0.1518NANormalDilated CMCM
2732Male0.333664NormalDilated CMCM
2867FemaleNA72NPRestrictive CMAtrial arrhythmia
2930FemaleNANANAAnomalous coronaryRestrictive CMCM
3029Male1.5291NormalARVC; VTVentricular arrhythmia
3118Male<0.151NANormalDilated CMCM
3236FemaleNA11NANormalDilated CMCM
3329Female0.16666NormalRestrictive CMInfiltrative process
3470FemaleNANANAIsolated plaqueDilated CMCM
3571FemaleNANANANormalRestrictive CMCM, infiltrative process
3646Male<0.1517NAIsolated plaqueDilated CMCM
3738Female<0.1577NANormalDilated CMCM
3850Male<0.1510NANormalDilated CMCM
3934Male<0.1515NANormalDilated CMCM
4067Female<0.15NANADiffuse plaqueDilated CMCM
4133MaleNANANANormalRestrictive CMCM, pericardial disease
4260Male0.32NANANPDilated CMCM
4326FemaleNANANANPDilated CMCM
4462Male<0.15NANADiffuse plaqueDilated CMChest pain
4527Female<0.15NANANPDilated CMCHF
4659Female0.29181Diffuse plaqueRestrictive CMCM
4732Male0.636433NormalDilated CMCM
4869Female<0.15NANALesion >70%: PCIRestrictive CMCHF, infiltrative disease
4945Male<0.15NANAIsolated plaqueDilated CMCM
5032Male<0.15NANAIsolated plaqueDilated CMCM
5154FemaleNANANADiffuse plaqueDilated CM; systemic sarcoidosisCM
5278Male2.04NANADiffuse plaqueDilated CMVentricular arrhythmia
5323Male1.1738NANPRestrictive CMDyspnea, infiltrative disease
5481Female<0.15NANADiffuse plaqueDilated CMCHF
5539MaleNANANANPDilated CMCM
5655Female17.230NANormalDilated CMCM
5722Male28.61140NormalDilated CMChest pain
5849FemaleNANANANormalRestrictive vs pericardialPericardial disease
5968Female<0.15NANA3 Lesions >70%; CABGDilated CMChest pain
6072Male0.55NANANPRestrictive CMCM
6148Male<0.156NANormalDilated CMCM
6224Male27.933NANormalDilated CMCM
6341Male15.2615NormalDilated CMCM
6442Male22.4NANAIsolated plaqueDilated CMChest pain
6548Female1.67NANANormalDilated CMVentricular arrhythmia
6641Male34.9NANANormalDilated CMVentricular arrhythmia

The patients were referred for recent-onset (<6 months) cardiomyopathy. While we do not have a parallel registry for this patient subset, they represent an estimated 60% to 70% of the patients admitted for recent-onset, unexplained cardiomyopathy. None of the patients studied had an established diagnosis and none had cardiovascular symptoms for more than 6 months.

The CMR imaging findings (positive or negative) led to the performance of EMB for confirmation or further clarification. The reverse was also true, however, if the EMB was inconclusive or when the EMB was performed independently from the decision to perform CMR imaging.

The time interval between the EMB and CMR imaging procedures ranged from 0 to 153 days (mean±standard deviation [SD], 19.2±35.0 days), in 71% (47 of 66), 8 or fewer days separated the two procedures (Table I). The clinical indications for testing included etiologic assessment of nonischemic dilated cardiomyopathy (DCM) in 45 patients, restrictive cardiomyopathy (RCM) in 19, and arrhythmogenic right ventricular cardiomyopathy (ARVC) in two patients (Table I).

The mean±SD ejection fraction of the DCM group was 28%±6% and for the restrictive cardiomyopathy group 57%±5%. This retrospective study was approved by the institutional review board.

Endomyocardial Biopsy

Biopsies of the right ventricular septum were obtained using standard transjugular technique. For each biopsy, a minimum of three samples were removed from different sites and each sample was immediately placed in 10% formalin; the mean±SD of EMB samples obtained per patient was 4.4±0.8. The fixed samples were then processed for routine staining with hematoxylin-eosin, Masson’s trichrome, Congo red, and Prussian blue.

Two of the authors with experience in cardiac histopathology (PBB and CVL) independently reviewed and scored all slides, blinded to clinical history, laboratory findings, and the CMR interpretation. The biopsy grade for each patient was qualitatively graded as 0=normal, 1=mild, 2=moderate, or 3=marked change for myocyte hypertrophy, myocyte degeneration, interstitial fibrosis, and fat. Inflammation was graded as myocarditis (definite), borderline (mild) myocarditis, and resolving myocarditis. All biopsies were also evaluated for the presence of amyloid and iron deposition. The final histologic grades for each category for each biopsy represent the mean of the two independent grades of the two reviewers. A grading concordance of ≤1 level of change or severity between the two reviewers occurred in >90% of the biopsy sets. Concordance levels >1 were resolved by consensus review.

CMR Imaging

All CMR studies conducted after September 1, 2004, were performed on a 1.5 Tesla scanner (Siemens Avanto, Erlangen, Germany). Studies prior to this date were performed on a 1.5 Tesla scanner (CVi, GE Healthcare, Milwaukee, WI). The standard evaluation for cardiomyopathy included multiplane cine imaging for ventricular size, mass, and function (left ventricular ejection fraction by Simpson’s rule), myocardial and liver T2 quantitation (using a multiecho gradient echo sequence for iron quantification),14 and LGE imaging. Cine imaging involved breathhold steady-state free precession and real-time in instances where breathholding or a rhythm precluded steady-state free precession cine imaging. LGE was typically performed 10 minutes after intravenous administration of 0.2 mmol/kg gadolinium-based contrast agent using a T1-weighted inversion-recovery gradient echo sequence,3 optimizing the inversion time for maximal contrast between nonenhancing and hyperenhancing myocardium. Patients with a creatinine clearance <30 mL/min were excluded from the study.

Two of the authors with expertise in CMR analysis and interpretation (JAD and SVR) independently reviewed all studies, blinded to the biopsy findings. Studies were evaluated for extent and pattern of hyperenhancement and wall motion analysis. Wall motion was graded as normal, hypokinetic, akinetic, and dyskinetic in a distribution of a 17-segment model. Hyperenhancement was graded as no hyperenhancement present, hyperenhancement of <50% of a segment or >50%, and demarcated endomyocardium, midmyocardium, and epimyocardium. Reviewers were to comment if the pattern of LGE was consistent with amyloid or another clinical finding and characterized as supportive of ischemic cardiomyopathy, nonischemic cardiomyopathy, presence of myocardial scar, infiltrative or inflammatory condition, hypertrophic cardiomyopathy, ARVC, systolic dysfunction with or without segmental wall motion abnormality. Scoring for wall motion and LGE was performed and any discrepancies in interpretation were subsequently resolved by consensus review. Discrepancies warranting consensus review occurred in <10% of studies.

Statistical Analysis

Because of the modest numbers of patients in various sets and subsets, the data are largely presented in descriptive and tabular format.

Results

The CMR-LGE images were nondiagnostic in 8 of 66 patients due to use of a nonoptimized protocol prior to September 2004 and one patient with claustrophobia required early termination (inadequate in 9 of 66 [14%] in 4 patients with DCM and 5 with RCM) (Table II, Table III, and Table IV). EMB tissue samples were inadequate (<3 independent interpretable fields of myocardium) in two patients (3%, both with DCM). No complications occurred with CMR. EMB led to pericardial tamponade in one patient (patient 65 [1.5%]) with extensive fatty infiltration and possible ARVC by EMB.

Table II.   Major Laboratory Findings by CMR and EMB
Patient No.LVEDV, mLLVESV, mLLVEF, %Mass, gCMR FindingsEMB FindingsDays Between CMR and EMB (Initial Study)
  1. Abbreviations: ARVC, arrhythmogenic right ventricular cardiomyopathy; CM, cardiomyopathy; CMR, cardiac magnetic resonance imaging; DME, delayed myocardial enhancement; EF, ejection fraction; EMB, endomyocardial biopsy; fat, adipose tissue; GE, General Electric; LV, left ventricular; LVEF, left ventricular ejection fraction; LVEDV, left ventricular end-diastolic volume; LVESV, left ventricular end-systolic volume; NA, not available; QNS, quantity not specified; WMA, wall motion abnormality; WNL, within normal limits.

1NANA45227Hypertrophic CMModerate myocyte hypertrophy, moderate fibrosis7 (CMR)
214812714396ScarMild myocyte hypertrophy15 (CMR)
3813063NANondiagnosticMild myocyte hypertrophy4 (EMB)
4NANANANAWNLMild myocyte hypertrophy0
5NANANANAInflammation or infiltrationBorderline myocarditis, moderate myocyte hypertrophy, moderate fibrosis4 (CMR)
61067133103Low EF and WMAMild myocyte hypertrophy0
725213546NALow EF and WMAModerate fibrosis2 (CMR)
81394965168WNLModerate myocyte hypertrophy32 (CMR)
91176346182AmyloidosisAmyloidosis, mild myocyte hypertrophy3 (CMR)
101701549220AmyloidosisAmyloidosis, moderate myocyte hypertrophy74 (CMR)
11683253NAWNLModerate myocyte hypertrophy89 (CMR)
12NANANANAFocal DMENormal26 (CMR)
13NANANANAWNLModerate myocyte hypertrophy, mild fibrosis8 (EMB)
1439734912149ScarMarked myocyte hypertrophy, mild fibrosis1 (EMB)
151659045143Nondiagnostic (GE)Moderate fat, mild fibrosis1 (EMB)
16124684576Low EF and WMAMarked myocyte hypertrophy, mild fibrosis8 (CMR)
171751252991Ischemic changesModerate myocyte hypertrophy, marked fibrosis98 (EMB)
18NANANANAScarMild myocyte hypertrophy, mild fibrosis1 (EMB)
1965354656Pericardial cystNormal0
20NANANANAInflammation or infiltrationGiant cell myocarditis, moderate fibrosis, mild myocyte hypertrophy1 (CMR)
211074261NAHyperdynamic LVAmyloidosis4 (CMR)
222422286138Nondiagnostic (GE)Resolving myocarditis, moderate myocyte hypertrophy, marked fibrosis48 (CMR)
23110337062Nondiagnostic (GE)Normal1 (CMR)
241478542NAInflammation or infiltration; right ventricular hyperenhance, possible ARVCModerate myocyte hypertrophy15 (CMR)
25NANANANANondiagnostic (GE)Mild myocyte hypertrophy35 (EMB)
263643339249ScarModerate myocyte hypertrophy5 (CMR)
271064260105WNLModerate myocyte hypertrophy, mild fibrosis0
281227638264Hypertrophic CMAmyloidosis, marked myocyte hypertrophy33 (CMR)
29NANANANAWNLMild myocyte hypertrophy33 (CMR)
302029254216ARVCModerate myocyte hypertrophy, mild fibrosis3 (CMR)
3126423810141ScarMarked myocyte hypertrophy, moderate fibrosis3 (CMR)
321166346145Septal bounceModerate myocyte hypertrophy, mild fibrosis5 (EMB)
3321012540184Nondiagnostic (GE)Marked myocyte hypertrophy, mild fibrosis4 (EMB)
341216050233Hypertrophic CMMild myocyte hypertrophy43 (CMR)
35NANANANAWNLBorderline myocarditis, mild fibrosis47 (CMR)
3625016733394ScarModerate myocyte hypertrophy, moderate fibrosis2 (CMR)
3713210520NALow EF and WMAQNS4 (CMR)
3818816910112ScarModerate myocyte hypertrophy, marked fibrosis0
3920016518NAScarModerate myocyte hypertrophy4 (EMB)
401689643132Reduced EFResolving myocarditis, moderate myocyte hypertrophy, marked fibrosis3 (CMR)
4179533355Nondiagnostic (GE)Marked myocyte hypertrophy, marked fibrosis8 (EMB)
42NANA10NAInflammation or infiltration (possible amyloid)Marked myocyte hypertrophy, moderate fibrosis124 (CMR)
431107631NAScarBorderline myocarditis, mild myocyte hypertrophy, moderate fibrosis2 (CMR)
441253274112WNLBorderline myocarditis, moderate myocyte hypertrophy, mild fibrosis28 (CMR)
45985841NAMild reduction EFModerate myocyte hypertrophy, mild fibrosis153 (EMB)
46NANA60NAWMAAmyloidosis, mild myocyte hypertrophy, moderate fibrosis2 (CMR)
473052857NAScarQNS8 (CMR)
48NANANANANondiagnostic (GE)Marked myocyte hypertrophy3 (CMR)
4920016020NAScarMarked myocyte hypertrophy1 (CMR)
5021717619245Reduced EFModerate myocyte hypertrophy66 (CMR)
5122016226130ScarMarked myocyte hypertrophy0
5298504988IschemicModerate myocyte hypertrophy, mild fibrosis, moderate fat150 (CMR)
531768552237Inflammation or infiltrationModerate myocyte hypertrophy, moderate fibrosis1 (CMR)
541008812165Takotsubo CMMild myocyte hypertrophy, moderate fibrosis3 (CMR)
551154527151Low EF and WMAModerate myocyte hypertrophy, mild fibrosis3 (CMR)
5695534497ScarMild fibrosis0
571296550138Inflammation or infiltrationBorderline myocarditis, moderate myocyte hypertrophy, moderate fibrosis0
58106356766WNLMild fibrosis75 (CMR)
59NANA10NAScarMarked myocyte hypertrophy3 (CMR)
60NANANANALimited studyAmyloidosis, marked myocyte hypertrophy, marked fibrosis3 (CMR)
6122620211160Reduced EFMarked myocyte hypertrophy, moderate fibrosis3 (EMB)
6219011142204Inflammation or infiltrationMyocarditis, mild myocyte hypertrophy, marked fibrosis1 (CMR)
631114361143Inflammation or infiltrationMyocarditis, mild myocarditis, moderate fibrosis1 (EMB)
641418341NAInflammation or infiltrationNormal1 (CMR)
65742466NAWNLMild fibrosis, marked fat, possible ARVC5 (EMB)
661467251NAInflammation or infiltrationBorderline myocarditis, moderate myocyte hypertrophy, mild fibrosis2 (CMR)
Table III.   Information Used for Cardiac Diagnosis
 CMR ImagingEMB
  1. Abbreviations: ARVC, arrhythmogenic right ventricular cardiomyopathy; CMR, cardiac magnetic resonance; EMB, endomyocardial biopsy; LV, left ventricular; NA, not applicable.

Dilated cardiomyopathy (n=45)
LV volumes, mass or function40 (89%)NA
Myocardial findings
 Ischemic3 (1 Takotsubo)NA
 Scar-dense fibrosis14 (31%)NA
 Histologic fibrosisNA28
  MildNA12
  Moderate to markedNA16
 Inflammation810
  Myocarditis8 (2 with negative EMB)10
   Borderline35
   Definite33 (1 giant cell)
   Resolving02
  Exclusions1 (systemic sarcoid)1 (systemic sarcoid)
 Cellular
  HypertrophyNA34
   MildNA10
   Moderate to markedNA24
 Normal myocardium43
 Inadequate sampleNA2
Nondiagnostic CMR4NA
Additional findings
ARVC21 (possible, >80% fat on EMB)
Complications01 (cardiac tamponade)
Table IV.   Information Used for Cardiac Diagnosis
 CMR ImagingEMB
  1. Abbreviations: ARVC, arrhythmogenic right ventricular cardiomyopathy; CM, cardiomyopathy; CMR, cardiac magnetic resonance; EMB, endomyocardial biopsy; LV, left ventricular; NA, not applicable.

Restrictive cardiomyopathy (n=19)
LV volumes, mass or function14 (74%)NA
Myocardial findings
 Ischemic0NA
 Scar-dense fibrosis0NA
 Histologic fibrosisNA10
  MildNA5
  Moderate to markedNA5
 Infiltration36
  Amyloidosis26
 Inflammation01
  Myocarditis01
   Borderline01
   Definite00
   Resolving00
   Exclusions1 (pulmonary sarcoid)1 (pulmonary sarcoid)
 Cellular
  HypertrophyNA15
   MildNA5
   Moderate to markedNA10
 Normal myocardium71
 Inadequate sampleNA0
Nondiagnostic CMR5NA
Additional findings30
 Pericardial cyst10
 Hypertrophic CM20
 Complications00

DCM and ARVC

CMR provided information on cardiac anatomy, chamber volumes, left ventricular size and mass, and systolic function (ejection fraction) in 40 of 45 patients with DCM (89%), extensive fibrosis scar in 14 (31%), and myocardial ischemia in 3 (7%), one of which had previously undiagnosed Takotsubo (stress) cardiomyopathy (patient No. 54) (Table II and Table III). ARVC was established by CMR in two patients (patients 24 and 30).

The EMB provided none of this information but was revealing at a histologic level. Cardiomyocyte hypertrophy was noted in 34 patients (76%), moderate to marked in 24 patients (53%), and interstitial fibrosis was noted in 28 (62%) patients, moderate to marked in 16 patients (36%). Based on the finding of extensive fat infiltration (>80%) in the 7 biopsy samples of patient No. 65, the diagnosis of ARVC was suspected but not confirmed by CMR.

Inflammatory Cardiomyopathy/Myocarditis  The diagnosis of myocarditis was independently rendered by both CMR imaging and EMB in 6 patients; 3 with definite myocarditis and 3 with borderline (mild) myocarditis by EMB. The diagnosis of giant cell myocarditis was established by EMB in one of these patients (No. 20) who had recurrent refractory ventricular tachycardia. CMR imaging raised the possibility of myocarditis in two additional patients but was not seen on EMB; the histology was normal in one patient (No. 64), whose procedures were separated by 1 day and as hypertrophy and fibrosis in another (No. 42) whose procedures had a gap of 124 days, suggesting that the CMR suspicion of myocarditis may have been correct when performed. EMB was read as borderline (mild) or resolving myocarditis in 4 additional patients, none of which were detected by CMR imaging. Neither procedure found evidence of myocardial inflammation in a patient with systemic sarcoidosis (No. 51).

Restrictive Cardiomyopathy

CMR imaging provided data on cardiac structure, left ventricular volume and mass, and systolic function in 14 (74%) patients with RCM. Hypertrophic cardiomyopathy was established by CMR imaging in 3 patients, one of which was found to have amyloid heart disease (patient No. 28) by EMB. CMR imaging revealed a previously undetected pericardial cyst in patient No. 19 to account for the clinical findings of impaired right ventricular filling with normal myocardium by EMB. Normal myocardium was noted by CMR imaging in 7 patients with a clinical picture of RCM.

EMB showed that 10 (53%) of the 19 patients with RCM had increased interstitial fibrosis (5 with moderate to marked) and 15 had cardiomyocyte hypertrophy (79%) (10 with moderate to marked). Normal myocardial histology was found in only 1 patient (patient No. 19 with CMR-detected pericardial cyst).

Amyloid Heart Disease  EMB identified amyloid heart disease in 6 (32%) of the 19 patients with RCM. Only 2 of the 6 (patients No. 9 and 10) were noted to have amyloid heart disease by CMR. Of the remaining 4 patients with EMB-proven amyloid infiltration, CMR imaging identified hypertrophic cardiomyopathy in 1 (No. 28), hyperdynamic systolic function in 1 (No. 21), and septal dyskinesia in 1 (No. 46) and was nondiagnostic in 1 (No. 60). Infiltrative disease was suspected by CMR imaging in patient No. 53, but was found to be cardiomyocyte hypertrophy and interstitial fibrosis by EMB.

Borderline myocarditis with mild interstitial fibrosis was found by EMB in 1 patient (No. 35) with normal myocardium by CMR. Evidence of inflammation (sarcoid heart disease) was not found by CMR or EMB in patient No. 11 who had pulmonic sarcoidosis and clinical RCM.

Additional Laboratory Information Used in the Evaluation of Cardiomyopathy

Fifty (76%) of the patients underwent coronary angiography: by catheterization in 49 and by computed tomographic angiography in 1 (Table I). Patient No. 59 with DCM had high grade 3 vessel occlusive coronary artery disease and underwent coronary artery bypass surgery, and patient No. 48 with RCM was found to have a >70% obstructive atherosclerotic lesion treated with angioplasty and stent deployment. A patient with RCM (No. 29) had an incidental anomalous coronary artery. The remaining 47 patients studied had normal coronary angiographic findings (n=29) or nonobstructive (<50%) plaque formation (n=18).

For the 11 patients with definite, borderline, or resolving myocarditis by EMB, 7 of 9 had elevated serum troponin I values (>0.15 ng/mL), 3 of 6 had elevated erythrocyte sedimentation rate (>20 mm/h), and 4 of 4 had elevated high-sensitivity C-reactive protein (hs-CRP) (>3.0 mg/L). However, for the 3 patients with definite myocarditis, 3 of 3 had elevated troponin, 3 of 3 had elevated sedimentation rate, and 2 of 2 had elevated hs-CRP. For patients with borderline (mild) myocarditis, the respective values were 3 of 4, 0 of 2, and 1 of 1 and for resolving myocarditis 1 of 2, 0 of 1, and 1 of 1.

Forty-five patients remained after excluding those with myocarditis (n=11) and amyloid (n=6) by EMB and those with occlusive coronary artery disease (n=2) by angiography of the 64 patients with adequate biopsy samples. In this group of 45 patients, 15 of 36 (42%) had elevated serum troponin, 13 of 24 (54%) had elevated sedimentation rate, and 9 of 12 (75%) had elevated hs-CRP.

Discussion

This study supports the concept that at the present time, CMR imaging and EMB are complementary procedures, each providing distinct information, in the evaluation of nonischemic cardiomyopathy.

In contrast to EMB, CMR imaging in heart failure provides important information on cardiac and vascular anatomy, pericardial conditions, size of cardiac chambers, myocardial hypertrophy, replacement fibrosis scar, and ventricular function. In addition, CMR imaging can uncover a number of other unsuspected anatomic-pathologic derangements. Examples from this study include ARVC (patients No. 24 and No. 30), pericardial cyst (No. 19), Takotsubo (stress) cardiomyopathy (No. 54), and hypertrophic cardiomyopathy (No. 1, 28). EMB by itself is not capable of providing any of this information, but when EMB is accompanied by a right heart catheterization, it can provide hemodynamic data, namely right heart, pulmonary artery and pulmonary artery occlusive (wedge) pressures, cardiac output, and right-sided blood oxygen saturations.

In contrast to CMR imaging, EMB provides details of myocardial histology in terms of myocyte size and hypertrophy and interstitial fibrosis. Degenerative change in cardiomyocytes can also be seen on EMB, but the rather subjective nature of its determination and analysis and the uncertainty of interpretation dissuaded the authors from submitting this finding. Aside from simply adding histologic findings to the patient’s cardiomyopathic condition and perhaps contributing to clinical impressions (eg, myocyte hypertrophy implies some chronicity, extensive interstitial fibrosis could account for diastolic dysfunction), one may question the clinical merit of determining the myocardial histology by EMB. There are currently no therapies to specifically reverse pathologic cardiomyocyte hypertrophy per se or interstitial fibrosis. The hemodynamic information obtained from the accompanying right heart catheterization can be useful in many of these patients.

The interface of these two procedures in the evaluation of nonischemic cardiomyopathy is most obvious in conditions of myocardial inflammation and infiltrative disease. A few of these conditions now appear to reside in the diagnostic domain of CMR; for example, iron-overload heart disease, ventricular fibrofatty dysplasia, and replacement fibrosis scar.1,2,4–7,14,15

This study does not support CMR imaging entirely replacing EMB in the diagnosis of amyloid heart disease. Of the 6 patients with EMB-proven amyloid infiltration of the heart, CMR revealed this condition in 2 patients (patients No. 9 and No. 10, true positives) and raised the question of amyloid deposition in patient No. 42, but not confirmed by EMB (a false-positive result for CMR). CMR did not uncover this condition in the 4 additional patients with EMB-proven amyloid infiltration (CMR false-negatives); one of these false-negative readings was due to the lack of LGE data related to patient intolerance (claustrophic), while the individual CMR interpretation in the remaining 3 patients was hypertrophic cardiomyopathy, hyperdynamic left ventricular, and a septal motion abnormality. The CMR diagnosis of cardiac amyloidosis is largely based on anatomic findings, gadolinium kinetics, and pattern of LGE and, thus, may be limited in its ability to convincingly diagnose early or mild stages of this condition. The greater diagnostic power of CMR in recent reports is likely related to studies performed in select patients with already established systemic or cardiac amyloidosis (ie, known amyloidosis) and to the continued refinement of CMR methodology.16–21 Despite the findings of the current study, the use of CMR imaging in diagnosing amyloid heart disease will continue to expand as methodology improves, particularly in patients with known systemic amyloid disease or notable clinical and/or echocardiographic features of amyloidosis.16,17,20 The role of EMB may then move to select patients with RCM.

A similar discussion unfolds for myocarditis. The overall prevalence of proven myocarditis by EMB, namely 4.7% (3 of 64) for definite and 17.2% (11 of 64) for combined definite, borderline (mild), and resolving myocarditis in this study, is within the range of those of most reported series.12,22–25 The CMR in the current study reported inflammation in 8 patients with DCM; 6 were confirmed by EMB (true-positives for CMR) as definite in 3 (1 with giant cell myocarditis) and borderline (mild) myocarditis in 3. Two patients with borderline myocarditis and 2 with resolving myocarditis via EMB were not detected by CMR imaging (false-negatives). Two patients with CMR-suspected inflammation were not confirmed as such by EMB, although the diagnostic studies were separated by 124 days in one of the patients (No. 42). One patient with RCM had borderline (mild) myocarditis by EMB alone (No. 35). Incidentally, EMB is currently the only method available to establish the diagnosis of giant cell myocarditis.

Much of the apparent disparity in the 2 techniques for diagnosing myocarditis in the current study is likely related to the limitations of each, namely the questionable ability of CMR imaging to detect borderline (mild) or resolving myocarditis and the imperfect sampling and somewhat subjective histologic interpretation of EMB (eg, “borderline” myocarditis).1,9–13,23,26 Importantly, CMR detected an LGE pattern consistent with myocarditis in the 3 patients with definite myocarditis by EMB, and determining whether a patient has borderline or resolving myocarditis may not alter management at the present time. Although the number of patients with myocarditis in the current study is relatively small, the results of CMR in the detection of this condition are consistent with published reports.11–13,26–29 DeCobelli and colleagues28 found CMR imaging to be diagnostic in 84% of patients with definite myocarditis and 44% in borderline cases. Again, the diagnostic yield and accuracy of CMR imaging is likely to increase with technological advancements and with the uniform application and refinement of the “Lake Louise criteria” for the CMR diagnosis of myocardial inflammation.27

The roles of CMR imaging and EMB have changed considerably during the course of this study. At the present time, CMR is indicated for most patients with unexplained dilated or restrictive cardiomyopathy. Renal dysfunction and implanted devices are limitations to this approach. For these patients, EMB may play a role in their diagnoses when inflammation or infiltrative disease is possible and clarification of such makes a therapeutic and prognostic difference. The same holds true when the results of CMR imaging are inconclusive or perhaps during the early stages of amyloid heart disease.

Markers of cardiac injury and inflammation have been applied over the years to the diagnosis of myocarditis with perhaps some increase in diagnostic yield.30,31 Both serum troponin and sedimentation rate were elevated in all 3 patients with definite myocarditis and hs-CRP was elevated in the 2 of these patients in whom it was measured. In borderline or resolving myocarditis by EMB, the yield of serum troponin dropped to 4 of 6 and sedimentation rate to 0 of 3, but remained at 2 of 2 for hs-CRP. The impact of these findings is greatly diminished by their high occurrence rates in this study in patients without myocarditis, amyloidosis, and occlusive coronary artery disease; 42% of the patients in this group had elevated serum troponin, 54% had an elevated sedimentation rate, and 75% had elevated hs-CRP. The high occurrence rates of these markers in nonmyocarditis forms of cardiomyopathy restrains their routine use as established discriminatory indicators for inflammatory heart disease.

Study Limitations

This study has a number of major limitations. It is a retrospective investigation with all the limitations inherent in this approach. The number of patients with any specific diagnosis (eg, amyloid heart disease, myocarditis) is relatively small. We can not exclude selection bias because this series included only patients with nonischemic cardiomyopathy who underwent both CMR imaging and EMB. The 2 procedures were not performed on the same day for most patients in this report, although in 71%, they were obtained within 8 days of each other. The earlier CMR studies (8 of 66 patients, 12%) were not performed using the most current technology and software and not all were accompanied by LGE; however, these units and methods are still in use at some centers.

Contemporary CMR now affords a broad armamentarium of tissue characterization techniques that are sufficiently robust for routine incorporation in cardiomyopathy and ARVC evaluations. While at the time that many of these studies were performed we did not perform rapid fat-suppressed cine imaging, T1 mapping and T2 mapping techniques, the results of our work support further prospective studies incorporating such techniques for consistent myocardial characterization irrespective of cardiac rhythm and breathholding capability of the patient. Patients with renal insufficiency were not studied. As discussed, the EMB is limited by random sampling of nonuniform cardiac pathology and some subjectivity in interpretation, although interobserver concordance was good for this study. Newer immunohistochemical and PCR analyses for viral genomes would likely have increased the diagnostic yield of viral-induced myocarditis in EMB samples.12,32 EMB sampling was only performed along the right ventricular septum; biopsies from both ventricles may have increased the diagnostic yield.33 With respect to the biomarkers, the number of patients tested was relatively small and so the interpretation of these is limited.

Conclusions

At present, CMR and EMB are still complementary, each providing unique information, in their general evaluation of nonischemic cardiomyopathy and its causes and consequences.

Ancillary