Author + information
- Received September 3, 2013
- Revision received October 8, 2013
- Accepted October 10, 2013
- Published online March 1, 2014.
- Marco Francone, MD, PhD∗,
- Cristina Chimenti, MD, PhD†,‡,
- Nicola Galea, MD∗,
- Fernanda Scopelliti, PhD§,
- Romina Verardo, PhD§,
- Roberto Galea, MD‖,
- Iacopo Carbone, MD∗,
- Carlo Catalano, MD∗,
- Francesco Fedele, MD† and
- Andrea Frustaci, MD†,§∗ ()
- ∗Department of Radiology, Oncology and Pathology, La Sapienza University, Rome, Italy
- †Department of Cardiovascular, Respiratory, Nefrologic, Anestesiologic and Geriatric Sciences, La Sapienza University, Rome, Italy
- ‡IRCCS San Raffaele La Pisana, Rome, Italy
- §IRCCS L. Spallanzani, Rome, Italy
- ‖Fondazione IRCSS Ospedale Maggiore Policlinico Università degli Studi di Milano, Rome, Italy
- ↵∗Reprint requests and correspondence:
Dr. Andrea Frustaci, Department of Cardiovascular Respiratory, Nefrologic, Anestesiologic and Geriatric Sciences, La Sapienza University, Viale del Policlinico 155, 00161 Rome, Italy.
Objectives The aim of this study was to determine whether clinical presentation and type of cell death in acute myocarditis might contribute to cardiac magnetic resonance (CMR) sensitivity.
Background Growing evidence indicates CMR is the reference noninvasive tool for the diagnosis of acute myocarditis. However, factors affecting CMR sensitivity are still unclear.
Methods We retrospectively evaluated 57 consecutive patients with a diagnosis of acute myocarditis made on the basis of clinical history (≤3 months) and endomyocardial biopsy evidence of lymphocytic infiltrates (≥14 infiltrating leukocytes/mm2 at immunohistochemistry) in association with damage of the adjacent myocytes and absence or minimal evidence of myocardial fibrosis. CMR acquisition protocol included T2-weighted (edema), early (hyperemia), and late (fibrosis/necrosis) gadolinium enhancement sequences. Presence of ≥2 CMR criteria denoted myocarditis. Type of cell death was evaluated by using in situ ligation with hairpin probes.
Results Three clinical myocarditis patterns were recognized: infarct-like (pattern 1, n = 21), cardiomyopathic (pattern 2, n = 21), and arrhythmic (pattern 3, n = 15). Tissue edema was observed in 81% of pattern 1, 28% of pattern 2, and 27% of pattern 3. Early enhancement was evident in 71% of pattern 1, 67% of pattern 2, and 40% of pattern 3. Late gadolinium enhancement was documented in 71% of pattern 1, 57% of pattern 2, and 47% of pattern 3. CMR sensitivity was significantly higher in pattern 1 (80%) compared with pattern 2 (57%) and pattern 3 (40%) (p < 0.05). Cell necrosis was the prevalent mechanism of death in pattern 1 compared with pattern 2 (p < 0.001) and pattern 3 (p < 0.05), whereas apoptosis prevailed in pattern 2 (p < 0.001 vs. pattern 1 and p < 0.05 vs. pattern 3).
Conclusions In acute myocarditis, CMR sensitivity is high for infarct-like, low for cardiomyopathic, and very low for arrhythmic clinical presentation; it correlates with the extent of cell necrosis–promoting expansion of interstitial space.
Clinical diagnosis of acute myocarditis is often challenging due to the wide spectrum of its presentations, ranging from a subclinical disease characterized by flu-like symptoms to an “infarct-like” syndrome with acute chest pain or to sudden death related to the new onset of arrhythmias and complete heart block (1). The true incidence of this condition is therefore unknown and certainly underestimated in the community. The diagnosis is usually presumptive, and serologic test results and conventional imaging tools such as echocardiography or selective coronary angiography fail to provide a definite diagnosis in most of the cases (1). For these reasons, endomyocardial biopsy (EMB), with use of the Dallas criteria, immunohistochemistry, and polymerase chain reaction for viral genomes, is still regarded as the diagnostic gold standard technique providing additional insight about possible underlying etiologies and pathogenic mechanisms (2).
Establishing a correct diagnosis of the disease may be crucial for a tailored therapeutic strategy (immunosuppressive, immunomodulatory, and/or antiviral) to reduce risk of progression to chronic active disease and dilated cardiomyopathy (3). However, widespread utilization of the EMB is limited in clinical practice by its invasiveness and by the possible associated inherent procedural risks; its use in acute myocarditis has been indicated when functional impairment, malignant arrhythmias, or failure of supportive treatment occur (4).
In this clinical setting, cardiac magnetic resonance (CMR) has become the reference noninvasive diagnostic tool allowing identification of the various hallmarks of myocardial inflammation represented by edema, fibrosis, and hyperemia and combining its unique tissue characterization capabilities with assessment of biventricular regional and global function (5–8).
Clinical experience, however, has shown that CMR signal abnormalities are variably observed in acute myocarditis, and sensitivity of the examination may be influenced by the pattern and extent of myocardial involvement, which actually reflects clinical presentation of disease. In the present article, we report CMR sensitivity in 57 patients with biopsy-proven acute myocarditis and its relationship with type of clinical presentation and mechanism of cell death.
We retrospectively reviewed our clinical database of 118 patients admitted to our institution during a 3-year period (from March 2008 to April 2011) with a clinical suspicion of acute myocarditis (clinical history ≤3 months) who underwent both CMR and EMB. Fifty-seven patients (48%) had a histological diagnosis of acute myocarditis based on histology and immunohistochemistry and represented our patient population. In particular, the histologic diagnosis of acute myocarditis included evidence of lymphocytic infiltrates in association with damage of the adjacent myocytes, according to the Dallas criteria (9) implemented by the immunohistochemical definition of inflammatory cells (≥14 infiltrating leukocytes/mm2, preferably T lymphocytes or activated T cells) (10) and the absence or minimal evidence of myocardial fibrosis, suggesting an acute process.
Patients were excluded from the analysis if there was evidence of significant coronary artery disease or valvular abnormalities, systemic diseases, and a clinical history ≥3 months. The local ethics committee provided approval for this investigation, and the patients gave their written informed consent.
CMR acquisition protocol
CMR imaging was performed with a 1.5-T system (Magnetom Avanto, Siemens Medical Systems, Erlangen, Germany) using a body and phased array coil. The body coil was adopted for measurements of the relative signal intensities of the myocardium and skeletal muscle because of its more homogeneous signal.
Cine steady-state free precession (cine-SSFP) CMR images were acquired during breath-holds in the short-axis, 2-chamber, and 4-chamber planes; on short-axis images, the left ventricle was completely encompassed from the base to the apex, acquiring a total of 10 to 12 images. Cine-SSFP images were obtained using the following parameters: repetition time 51.3, echo time 1.21, a flip angle of 45°, an 8-mm slice thickness, a matrix of 256 × 256, a field of view ranging from 340 to 400 mm, and a voxel size of 2.0 × 1.3 × 8.0 mm.
For T2-weighted short-tau inversion recovery (T2w-STIR) imaging, a breath-hold black-blood, segmented turbo spin echo technique was adopted, using a triple inversion recovery preparation module (repetition time, 2 R-to-R intervals; echo time 75 ms; flip angle 180°; inversion time 170 ms; slice thickness 8 mm; no interslice gap; field of view 340 to 400 mm; matrix 256 × 256; and voxel size 2.3 × 1.3 × 8 mm). Technical details of this sequence have been described elsewhere (11).
The T2w-STIR images were acquired on short-axis planes covering the entire left ventricle during 6 to 8 consecutive breath-holds. Each slice was obtained during an end-expiratory breath-hold of 12 to 15 s, depending on the patient's heart rate. T1-weighted turbo spin echo images using a free-breathing non–electrocardiogram (ECG)-gated technique were also acquired in 4 identical axial slices both before and after intravenous administration of 0.1 mmol/kg of gadobenate dimeglumine (Gd-BOPTA) (Multihance, Bracco Imaging, Milan, Italy) at 2 ml/s; the sequence was started immediately after contrast injection (sequence duration 3 to 4 min).
Finally, an inversion recovery contrast-enhanced CMR (2-dimensional inversion recovery fast gradient echo technique) sequence was acquired on late (10 to 15 min) imaging after gadolinium injection (late gadolinium enhancement [LGE]), setting the acquisition window to mid-end diastole. Sequence parameters were as follows: repetition time 9.6/4.4; matrix 256 × 208; flip angle 25°; slice thickness 5.0 mm; and slice spacing 5.0 mm. Meticulous attention was paid to the inversion time to null normal myocardium, with typical values of 250 to 300 ms.
All CMR studies were analyzed off-line in consensus by 2 experienced observers (M.F. and I.C. [10 and 11 years of experience, respectively]) blinded to EMB results, using a workstation with dedicated cardiac software (Leonardo and Argus, Siemens Medical Systems).
For the evaluation of left ventricular (LV) global and regional function and calculation of LV mass, the endocardial and epicardial borders were manually drawn in the end-diastolic and end-systolic short-axis cine-SSFP images. Papillary muscles and trabeculatures were not included in the myocardium. LV end-diastolic volume, LV end-systolic volume, ejection fraction, and LV mass were determined.
Myocardial Edema Assessment: T2w-STIR
Presence of tissue edema was identified from T2w-STIR sequences using the T2-ratio method of quantification by manually outlining 2 separate regions of interest, respectively, within the entire LV myocardium and the visible skeletal muscle (serratus anterior, combination of teres minor and infraspinatus, subscapularis, or a combination of major and minor pectoralis depending on structure visibility and signal intensity homogeneity) (6,12). The myocardial signal intensity was related to that of the skeletal muscle to calculate the T2 ratio; a ratio ≥2 was considered to be positive and reflective of the presence of global myocardial edema (5).
Hyperemia and Fibrosis: Early and LGE Assessment
As reported in Bondarenko et al. (13), LGE was considered positive if the signal intensity was >5 SDs above the remote myocardium and presented a nonischemic pattern of distribution (i.e., subepicardial or mesocardial enhancement).
As another criterion for tissue inflammation, early enhancement reflecting hyperemia and capillary leakage was assessed by delineating endocardial and epicardial contours in T1 images, which were acquired before and over the first 3 min after contrast injection. A region of interest was traced within normally appearing skeletal muscle. All contours were traced in pre-contrast images and copied onto post-contrast images. The myocardial enhancement (percent signal intensity increase post-contrast versus pre-contrast image) was normalized to the enhancement of the skeletal muscle early enhancement ratio (EER); if the signal intensity ratio was ≥4, CMR was considered positive for inflammation (14).
EMB and immunohistochemistry
EMB were performed in the septal-apical region of the left or biventricular chamber, which were approached by a 7-F (501-613A, Cordis Corporation, Miami, Florida) long sheet and identified on a radiographic view by using flashing of contrast medium. Among the 118 patients, 83 received a biventricular EMB (70%) and 35 (30%) received an LV EMB. Three to 6 right ventricular and/or 3 to 6 LV biopsy specimens (from 2 to 5 mm3) were drawn from each patient, on the basis of the clinical condition and the need of tissue samples.
Three to 5 samples per each ventricular chamber were drawn, processed for histology and immunohistochemistry, and read by a pathologist blinded to the clinical data, as previously described (15). Two to 3 samples per patient were snap-frozen in liquid nitrogen for molecular biology studies. Multiple 5-micron-thick sections were stained with hematoxylin-eosin, Miller's elastic Van Gieson, and Masson's trichrome, and examined by using light microscopy. In all patients, immunohistochemistry for the characterization of inflammatory infiltrates was performed by use of the following antibodies: CD45, CD43, CD45RO, CD20, CD68, CD4, and CD8.
The presence of ≥14 infiltrating leukocytes/mm2 and/or the presence of >2 CD3-positive lymphocytes per high-power field (400-fold magnification), often adherent to the contour of the cardiomyocytes, were considered diagnostic for active myocarditis.
Evaluation of cell death
In situ ligation of hairpin probes with single-base 3' overhangs (hairpin 1) or blunt ends (hairpin 2) were used to measure cardiomyocyte apoptosis and necrosis, respectively (16). Cardiomyocytes were labeled by using alpha-sarcomeric actin antibody staining (clone C5C, Sigma, St. Louis, Missouri). Nuclei were stained with 4',6-diamidino-2-phenylindole. Specimens were examined with a confocal microscope. Control specimens consisted of surgical LV samples from patients with mitral stenosis undergoing valve replacement who had normal LV function. These endomyocardial fragments, although not obtained from healthy individuals, were derived from a not-overloaded chamber and were considered the samples closest to normal endomyocardial tissue.
Molecular biology studies
Polymerase chain reaction and reverse-transcriptase polymerase chain reaction analysis was performed on 2 frozen EMB tissue samples to search for the most common DNA (adenovirus, cytomegalovirus, parvovirus B19, Epstein-Barr virus, human herpes virus 6, and herpes simplex virus 1 and 2) and RNA (enterovirus, influenza virus A and B, hepatitis C virus) cardiotropic viruses as described earlier (16).
Data were analyzed using SPSS version 15.0 (IBM SPSS Statistics, IBM Corporation, Armonk, New York). Normal distribution of variables was assessed using the Kolmogorov-Smirnov test. Quantitative measurements are expressed as mean ± SD, and categorical data are presented as absolute frequencies and percent values.
Significance of results was assessed by using a Wilcoxon-Mann-Whitney rank sum test when 2 different groups were directly compared; comparisons among all 3 groups were performed using the Kruskal-Wallis test. A post-hoc evaluation was also conducted using a Mann-Whitney analysis. A 2-tailed p value <0.05 was considered statistically significant.
Clinical presentation and EMB
Three distinctive patterns of clinical presentation were observed: 1) infarct-like pattern (pattern 1) including fever, chest pain, ST-segment elevation at ECG and increase of serum troponin I (n = 21 [13 male subjects]; mean age 44.7 ± 15.3 years); 2) cardiomyopathic pattern (pattern 2) characterized by LV dysfunction and severe heart failure (New York Heart Association functional class III to IV) in the absence of ECG, serologic, or systemic abnormalities (n = 21 [18 male subjects]; mean age 52.5 ± 16.2 years); and 3) arrhythmic pattern (pattern 3) consisting of sudden occurrence of life-threatening ventricular arrhythmias in the absence of systemic evidence of inflammation/infection (n = 15 [10 male subjects]; mean age 52.6 ± 9.0 years). No significant differences were found among groups in terms of age, sex, or history of a recent flu-like syndrome.
The T2 ratio was positive in 27 cases (47%) characterized by focal or diffuse distribution of edema observed, respectively, in 35% (n = 20) and 12% (n = 7) of the cases. EER was elevated in 61% (35 of 57), whereas LGE was present in 60% (34 of 57) patients with either a mid-wall or subepicardial pattern of distribution. Predominant involvement was observed in the mid-basal lateral segments (n = 21 cases) either isolated or with concomitant septal (n = 8 cases) or diffuse (n = 4 cases) enhancement.
When CMR findings were matched with patients' clinical presentations, the following results were observed. EER was positive in 71% (15 of 21) of cases with pattern 1, 67% (14 of 21) with pattern 2, and 40% (6 of 15) with pattern 3 (p = 0.133). Tissue edema was observed in 81% (17 of 21) of patients with pattern 1, 28% (6 of 21) with pattern 2, and 27% (4 of 15) with pattern 3, respectively (p = 0.001) (Figs. 1B, 2B, and 3B). LGE suggesting myocarditis (subepicardial, mid-wall, or diffuse) was documented in 71% of cases (15 of 21) with pattern 1, 57% (12 of 21) with pattern 2, and 47% (7 of 15) with pattern 3 (p = 0.314).
Overall, the CMR examination results were negative for myocarditis (i.e., no evidence of LGE, edema, or EER) in 19% (4 of 21) patients with pattern 1, 33% (7 of 21) with pattern 2, and 46% (7 of 15) with pattern 3; there were no statistically significant differences among groups (p = 0.21). The presence of at least 2 CMR-positive features (according to the Lake Louise diagnostic criteria) (6) was observed in 17 of 21 patients with pattern 1, 12 of 21 with pattern 2, and 6 of 15 with pattern 3, with a sensitivity of 80%, 57%, and 40% for each pattern, respectively (p = 0.042).
Global LV function was depressed and end-diastolic volume was increased in pattern 2 compared with the other 2 groups (p < 0.001) (Table 1). Diagnostic performances of each individual CMR technique per group are plotted in Figure 4.
Histological and molecular biology findings
In all patients, histological analysis revealed active myocarditis with diffuse inflammatory infiltrates associated with focal necrosis of the adjacent myocytes (meeting the Dallas criteria) with interstitial and focal replacement fibrosis in most specimens (Figs. 1C, 2C, and 3C). The infiltrates included mainly activated T cells (CD45RO+, CD3+).
Evaluation of cell death
Cardiomyocyte apoptotic and necrotic cell death were, respectively, 896-fold (patients 19,707 ± 20,345/106; control 22 ± 10/106 [p < 0.001]) and 29-fold (patients 7,700 ± 5,681/106; control 265 ± 97/106 [p < 0.001]) greater in patients with acute myocarditis than in control specimens.
Comparing different pattern of myocarditis, we observed that in patients with an ischemic pattern, cardiomyocyte necrosis was 2.9-fold higher than cardiomyocyte apoptosis (apoptosis 4,767 ± 4,954/106; necrosis 13,812 ± 6,068/106 [p < 0.050]); in patients with a cardiomyopathic pattern, cardiomyocyte apoptosis was 18.7-fold higher than necrosis (apoptosis 48,473 ± 21,373/106; necrosis 2,581 ± 5,770/106 [p < 0.001]). Patients with an arrhythmic pattern had similar levels of apoptosis and necrosis (apoptosis 5,881 ± 4,453/106; necrosis 6,708 ± 4,529/106 [p = 0.80]).
The comparison of cardiomyocyte necrosis in the 3 different patterns revealed a statistically significant difference between patterns 1 and 2 (p < 0.001) and patterns 1 and 3 (p < 0.05) (Fig. 4). The comparison of cardiomyocyte apoptosis among groups showed a statistically significant difference between patterns 2 and 1 (p < 0.001) and patterns 2 and 3 (p < 0.05) (Fig. 5).
Polymerase chain reaction analysis showed the presence of a myocardial viral infection in 10 subjects (17.5%) (adenovirus in 4, Epstein-Barr virus in 2, coinfection of adenovirus/Epstein-Barr in 1, and hepatitis C virus in 3). No differences were identified in terms of percentage of myocardial virus detection in the 3 patterns of myocarditis (19% in pattern 1, 19% in pattern 2, and 13% in pattern 3).
The present study analyzed CMR sensitivity in human acute myocarditis and correlated the type of clinical presentation with expansion of intercellular space reflecting different mechanisms of cell death. The results of our study provide evidence that sensitivity of CMR is elevated in patients with an infarct-like clinical presentation, moderate in patients with a cardiomyopathic pattern, and low in those with an arrhythmic clinical profile.
In our patient series, when a combination of at least 2 of 3 (i.e., edema and/or EER and/or LGE) positive CMR imaging features were applied for the diagnosis of disease (Lake Louise criteria), we found a sensitivity ranging from 80% in the infarct-like type to 40% in the arrhythmic pattern, with remarkable differences compared with the average 67% reported in the pooled data of the white paper published by Friedrich et al. (5). This heterogeneous sensitivity was also confirmed from the analysis of individual diagnostic contributions of different techniques, and was more pronounced for the presence of edema (ranging between 27% and 81%) and less evident for LGE (which varied between 47% and 71%).
These results suggest CMR users and clinicians should be cautious in the exclusion of an inflammatory process in the absence of conventional CMR diagnostic criteria, even in patients with acute myocarditis. As a consequence, whenever clinical manifestations are severe (i.e., LV deterioration, life-threatening arrhythmias, and/or unexplained syncope) and suggestive of acute myocarditis, negative CMR findings do not prohibit an invasive study, including EMB.
Mechanisms of CMR inflammatory findings
Our study findings likely reflect the pathological substrates observed in our specimens with variable expression of the 3 distinctive morphological signs of disease consisting of edema, hyperemia, and fibrosis/necrosis. They directly influence CMR's capability to detect related signal abnormalities.
Interstitial myocardial edema is part of the inflammatory response in acute myocarditis due to the accumulation of osmotically active substances within the interstitium combined with increased endothelial vasopermeability, and its presence has been reported in previous studies using a T2w-STIR technique with a sensitivity and specificity of 84% and 74%, respectively (6,16). CMR detection of edema is feasible with T2 sequences due to the transverse relaxation time increase associated with water accumulation and likely depends on the intercellular space expansion associated with active inflammation (17). The recent implementation of a T2-mapping technique has enabled an objective and quantitative detection of myocardial inflammation, relatively insensitive to cardiac and respiratory motion (18).
Inflammation-associated hyperemic phenomena are detectable on early enhancement images as regional or global areas of increased myocardial T1 signal intensity reflecting the increased vascular permeability and extracellular fluid space expansion. Tissue necrosis can be depicted on LGE as a consequence of pathological intracellular contrast uptake related to the sarcolemmal rupture that occurs in the presence of cardiomyocyte damage (6,19,20). Gadolinium chelates cannot penetrate into healthy or apoptotic cardiomyocytes due to the lack of membrane damage, resulting in lack of pathological LGE with obvious potentially intriguing implications related to our study results (21).
We started with these generally known and accepted facts regarding water and gadolinium kinetics and their relations with different types of myocardial damage. We then tried to speculate on potential mechanisms causing this remarkable sensitivity variation among our 3 patient groups combining type of cell death (causing a different expansion of the interstitial space) with CMR findings.
The infarct-like pattern generally represents the most typical and frequent initial manifestation of the disease, characterized by severe and/or recurring acute chest pain, ST-segment elevation, and a significant increase in serological markers of inflammation with a slight troponin increase. Remarkably, as observed in our series, these striking manifestations are usually associated with preserved cardiac dimensions and function (1,22).
In this group, we observed the highest sensitivity of CMR, ranging between 81% for T2w-STIR and 71% for both EER and LGE. The most common hypothesis advanced to explain this finding is the predominant “microvascular” involvement associated with this clinical onset of disease, causing enhanced diffusion of free-water in the intercellular space with related T2-relaxation time increase (22).
The extracellular space expansion and the focal cardiomyocyte necrosis may both cause increased gadolinium accumulation with subsequent hyperemia and late enhancement on EER and LGE techniques, respectively, appearing in the typical subepicardial/mid-basal LV lateral wall pattern (5,22). Indeed, in this group of patients, cell necrosis was the prevalent mechanism of cardiomyocyte death compared with the other patterns. These results confirm in humans the data obtained from experimental myocarditis (23), suggesting a major contribution of cell necrosis to expansion of interstitial space, promoting edema, inflammatory cell infiltration, and myocardial fibrosis. Interestingly, in 2 cases from this group, there was evidence of positive edema and EER with no signs of LGE. This finding might be attributable to the presence of mild, diffuse myocardial damage, which is poorly detectable with conventional inversion recovery sequences using a semiquantitative approach (although associated with typical acute inflammatory changes as explained earlier). As shown in promising initial results (24), a T1-mapping technique could overcome this weakness of CMR and provide further information in the presence of diffuse forms of disease.
The main clinical feature in this second group of patients was the presence of persisting dyspnea with associated symptoms of heart failure. Pathogenesis of ventricular dilation in acute myocarditis can be related to direct damage/dysfunction of the cardiomyocytes by the etiologic agent causing extensive myocardial injury (3,25) or mediated by abnormal T-cell reaction to viral-related or segregated myocardial antigens; this results in higher and prolonged disease activity and therefore ventricular dysfunction (22).
The low sensitivity of T2w-STIR sequences in this group (28%) may represent the morphological expression of a prolonged, subacute myocardial inflammation/infection with progressive water reabsorption and less evident systemic signs of inflammation. Thus, the most accurate approaches for inflammatory cardiomyopathy diagnosis are EER and LGE techniques; according to a previous report, the pattern of enhancement in this group was characterized by a predominant mid-wall septal involvement (22).
However, CMR sensitivity in these patients was lower compared with the previous pattern. This outcome can be justified by the major prevalence of cell apoptosis that notoriously preserves the integrity of cell membrane, limiting the expansion of the interstitial space, and thus reducing the CMR signal changes.
The clinical scenario of arrhythmic onset of disease mainly characterized by palpitations was the most difficult to diagnose with CMR, as suggested by the overall sensitivity of different techniques ranging between 40% for EER, 27% for T2w-STIR, and 47% for LGE imaging. Overall, 46% of patients with histological diagnosis of acute myocarditis had a negative finding on CMR examination.
Finding an explanation for the lower diagnostic performances in this subgroup is challenging and likely related to multifactorial aspects regarding either technical or pathological issues. Certainly, the presence of heart rate irregularity is a main limitation because arrhythmic patients are, by definition, prone to motion artifacts related to suboptimal ECG synchronization, with lower diagnostic quality of the images and subsequent impact on lesions detectability. From a pathological point of view, myocardial inflammation appeared in the last cohort to be more focal and likely involving the cardiac conduction tissue (26). Indeed, the low level of cell necrosis and the mild compromise of cardiac volume and function observed seem to support this interpretation. In this setting, the relatively low contrast and spatial resolution of CMR can hamper the diagnostic accuracy whenever the process is limited in terms of spatial extent and disease activity.
Comparison of our data with sensitivity values and overall diagnostic performances of the white paper results can be affected by a nonrigorous application of the suggested acquisition protocol. In particular, we used a slice thickness of 8 mm in T2w-STIR sequences instead of the recommended 10 to 15 mm (5), with possible reflections on the signal intensity of both myocardium and skeletal muscle. Nonetheless, use of T2-ratio quantification as a marker of edema would minimize this potential influence.
Second, the applied EER thresholds of 4.0 were previously validated using conventional gadolinium chelates (gadopentetate dimeglumine), whereas in our patient cohort, we administered a high relaxivity contrast agent (Gd-BOPTA) at a standard single dose of 0.1 mmol/kg (5). Our study results might have even overestimated the presence of inflammation-associated positive hyperemic phenomena provided by the EER measurements due to a more pronounced T1-shortening effect associated with Gd-BOPTA (27). This effect may have led to a proportional increase of myocardial enhancement already in the early phase after injection. Further focused clinical studies could eventually define adjusted thresholds of EER by using Gd-BOPTA.
LV biopsy specimens are taken from the subendocardial layer and may not represent a direct investigation of CMR-positive areas involving the subepicardial and mesocardial layers. Nevertheless, the data from experimental myocarditis show only a quantitative histological difference between outer and inner LV layers, and thus did not compromise the reliability of our results.
Furthermore, our retrospective analysis included only patients with a histological diagnosis of acute myocarditis. Therefore, no false-positive cases were available in our patient cohort precluding the assessment of CMR specificity.
In acute myocarditis, CMR sensitivity is high for infarct-like myocarditis, low for cardiomyopathy, and very low for arrhythmic clinical presentation. It correlates with the extent of cardiomyocyte necrosis promoting an expansion of interstitial space. EMB may be required in CMR-negative subjects with electrical instability and/or cardiac deterioration for a final diagnosis.
The authors thank Proffesor Italo Nofroni of Sapienza University of Rome and Proffesor Laura Recchia of the University of Molise for their precious and kind support in statistical analysis.
The study was supported by the Italian Ministry of Health (grants RF-2009-1511346, RBFR081CCS, and MRAR08Y012). All authors have reported that they have no relationships relevant to the contents of this paper to disclose. The first 2 authors contributed equally to this study.
- Abbreviations and Acronyms
- cine steady-state free precession
- cardiac magnetic resonance
- early enhancement ratio
- endomyocardial biopsy
- gadobenate dimeglumine
- late gadolinium enhancement
- left ventricular
- T2-weighted short-tau inversion recovery
- Received September 3, 2013.
- Revision received October 8, 2013.
- Accepted October 10, 2013.
- American College of Cardiology Foundation
- Chow L.H.,
- Radio S.J.,
- Sears T.D.,
- McManus B.M.
- Dennert R.,
- Crijns H.J.,
- Heymans S.
- Cooper L.T. Jr..,
- Mensah G.A.,
- Baddour L.M.,
- et al.
- Friedrich M.G.,
- Sechtem U.,
- Schulz-Menger J.,
- et al.
- Abdel-Aty H.,
- Boye P.,
- Zagrosek A.,
- et al.
- Pennell D.J.,
- Sechtem U.P.,
- Higgins C.B.,
- et al.
- Hundley W.G.,
- Bluemke D.A.,
- Finn J.P.,
- et al.
- Friedrich M.G.,
- Strohm O.,
- Schulz-Menger J.,
- Marciniak H.,
- Luft F.C.,
- Dietz R.
- Frustaci A.,
- Russo M.A.,
- Chimenti C.
- Chimenti C.,
- Kajstura J.,
- Torella D.,
- et al.
- Thavendiranathan P.,
- Walls M.,
- Giri S.,
- et al.
- Abdel-Aty H.,
- Zagrosek A.,
- Schulz-Menger J.,
- et al.
- Holmvang G.,
- Dec G.W.
- Mahrholdt H.,
- Wagner A.,
- Deluigi C.C.,
- et al.
- Ferreira V.M.,
- Piechnik S.K.,
- Dall'armellina E.,
- et al.
- Mason J.W.
- Frustaci A.,
- Chimenti C.