Abstract
BACKGROUND
Diagnosing myocarditis is challenged by nonspecific clinical signs and symptoms and low accuracy of endomyocardial biopsy. Cardiac magnetic resonance imaging (CMR) provides both cardiac anatomy and tissue characterization in this setting, but the prognostic value of this method as a primary assessment tool in patients with suspected myocarditis remains limited.
OBJECTIVES
This study sought to determine cardiac event-free survival of a consecutive cohort with suspected myocarditis with regard to CMR findings.
METHODS
Six hundred seventy patients with suspected myocarditis underwent CMR including late gadolinium enhancement (LGE) parameters between 2002 and 2015 and were included and followed. We performed multivariable model for major adverse cardiovascular events (MACE) and determined the continuous net reclassification improvement by LGE markers.
RESULTS
At a median follow-up of 4.7 years (interquartile range [IQR]: 2.3 to 7.3 years), 98 patients experienced a MACE. Two hundred ninety-four (44%) patients showed LGE presence, which was associated with a more than doubling risk of MACE (hazard ratio [HR]: 2.22; 95% confidence interval [CI]: 1.47 to 3.35; p < 0.001). Annualized MACE rates were 4.8% and 2.1% corresponding to LGE presence and absence, respectively (p < 0.001). In the multivariable model, LGE presence maintained significant association with MACE (HR: 1.72; 95% CI: 1.08 to 2.76; p = 0.023). The computed continuous net reclassification improvement was 0.39 (95% CI: 0.10 to 0.67) when LGE presence was added to the multivariable model for MACE. Regarding location and pattern, septal and midwall LGE showed strongest associations with MACE (HR: 2.55; 95% CI: 1.77 to 3.83 and HR: 2.39; 95% CI: 1.54 to 3.69, respectively; both p < 0.001). A patchy distribution portended to a near 3-fold increased hazard to MACE (HR: 2.93; 95% CI: 1.79 to 4.80; p < 0.001). LGE extent (per 10% increase) corresponded to a 79% increase in risk of MACE (HR: 1.79; 95% CI: 1.25 to 2.57; p = 0.002). A normal CMR study corresponded to low annual MACE and death rates of 0.8% and 0.3%, respectively.
CONCLUSIONS
CMR tissue characterization provides effective risk stratification in patients with suspected myocarditis.
Keywords: cardiovascular magnetic resonance imaging, CMR, extracellular volume, myocarditis, outcome
Myocarditis remains a leading cause of sudden cardiac death in athletes (1) and in dilated cardiomyopathy (2). Diagnosis of this condition is challenging given its nonspecific signs and symptoms (3), a lack of diagnostic reference standard, and low accuracies of electrocardiogram (ECG), echocardiography, biomarkers, and even invasive endomyocardial biopsy (EMB) (4,5). With its technical benefits of localizing regional dysfunction and matching tissue characterization, cardiac magnetic resonance imaging (CMR) has become the primary imaging tool in many centers characterizing disease severity and planning of patient management (6). The Lake Louise criteria had established the current recommended diagnostic criteria incorporating early gadolinium enhancement, late gadolinium enhancement (LGE), and T2-weighted imaging for edema detection (6). However, prognostic evidence using CMR for patients with suspected myocarditis currently remains limited. The present study aimed to evaluate the prognostic value of tissue characterization by CMR in risk stratification of patients presenting with suspected myocarditis.
SEE PAGE 1988
METHODS
STUDY POPULATION.
The study included consecutive patients referred by their treating physician to undergo CMR for “suspected myocarditis” as the primary clinical question between December 2002 and December 2015 at our center. We included patients with presenting signs/symptoms of any 1 of these 3 groups: 1) acute chest pain syndromes with symptom onset <2 weeks before CMR; 2) subacute (onset ≥2 weeks) of dyspnea or signs of left ventricular (LV) dysfunction; and 3) subacute (onset ≥2 weeks) presentation of ventricular arrhythmias syncopal spells or abnormal ECG.
Exclusion criteria included: 1) any evidence of coronary artery disease (CAD) by either previous documented medical history, any imaging findings of CAD, or significant epicardial coronary stenosis by invasive coronary angiography; 2) any evidence of hypertrophic cardiomyopathy, arrhythmogenic right ventricular cardiomyopathy, cardiac sarcoidosis, or cardiac amyloidosis; and 3) any evidence of Takotsubo cardiomyopathy, constrictive pericarditis, Loeffler endocarditis, ventricular noncompaction, cardiac tumor, pulmonary embolism, or severe valve disease. Clinical data, cardiac biomarkers, and ECG at the time of the CMR were analyzed. ECGs were analyzed according to the following ECG criteria: low QRS voltage was defined by ≤5 mm in limb leads or ≤10 mm in all precordial leads (7), QTc prolongation >450 ms for males and >470 ms for females (8), and ST-depression as ≥0.1 mV at 80 ms from the J point, asymmetrical T-wave inversion ≥0.1 mV deep in 2 or more leads except aVR (9), and Q-wave as >0.3 mV in depth and/or >40 ms in duration in at least 2 contiguous leads except aVR (10). T-wave inversion was defined by negative amplitude ≥1 mm in at least 2 contiguous leads (11) and ST-segment elevation after the J point in 2 contiguous leads with the cutoff points of ≥0.2 mV in men and ≥0.15 mV in women in leads V2–V3, and/or ≥0.1 mV in other leads (12).
CMR IMAGING PROTOCOL AND IMAGE POST-PROCESSING.
Patient scanning was performed with a 3.0-T or a 1.5-T system (Tim Trio and Aera, Siemens, Erlangen, Germany). A standardized CMR consisting of cine steady-state free precession imaging (repetition time [TR], 3.4 ms; echo time [TE], 1.2 ms; in-plane spatial resolution, 1.6 2 mm) for LV function and LV mass was used. Cine imaging was obtained in 8 to 14 matching short-axis (8-mm thick with no gap) and 3 radial long-axis planes. All patients underwent a fast-gradient echo technique LGE imaging protocol (TR, 4.8 ms; TE, 1.3 ms; inversion time, 200 to 300 ms) to detect fibrosis, using a segmented inversion-recovery pulse sequence starting 10 to 15 min after a weight-based injection (cumulative dose 0. 15 mmol/kg) of gadolinium diethylenetriamine penta-acetic acid (Magnevist, Bayer HealthCare Pharmaceuticals, Wayne, New Jersey) or gadobenate dimeglumine (Multihance, Bracco Diagnostic, Princeton, New Jersey). In patients with estimated glomerular filtration rate <60 ml/min/1.73 m2, contrast dose was restricted to 0.1 mmol/kg or 20 ml, depending on which was lower in volume based on our institutional policy (13). T1 measurements were acquired using a validated cine Look-Locker sequence (14), with a non–slice-selective adiabatic inversion pulse, followed by segmented gradient-echo acquisition for 17 cardiac phases/times after inversion (TE = 2.5 ms; TR = 5.5 ms; flip angle = 10°; 192 × 128 matrix; 6-mm slice), spread over 2 cardiac cycles (inversion time increments for T1 measurements of 100 ms pre-contrast, and 55 ms post-contrast, slice thickness 8 mm, TR >RR intervals pre-contrast and 3 RR intervals post-contrast). The Look-Locker sequence was performed in short-axis slices at basal, mid, and apical LV levels. T1 mapping images were acquired in the same LV short-axis slices, once before and up to 3 times after the injection of gadolinium spanning across a post-contrast period of approximately 30 min. Commercially available software (MASS v15, 2008, Medis, Leiden, the Netherlands) was used to post-process and quantify all CMR images. CMR variables were calculated. Further wall motion abnormalities (WMA), and pleural and pericardial effusion were visually assessed. Epicardial and endocardial contours were placed manually on all LGE images, then LGE mass was quantified by using the full width half maximum signal intensity threshold cutoff technique above the mean intensity of remote myocardium in the same slice (15) (Central Illustration), and calculation of LGE percentage according to the LV mass was performed. LGE was further assessed by its localization (anterior, inferior, septal, lateral), its distribution (linear, patchy, or diffuse), and its pattern (epicardial, midwall) (Figure 1).
CENTRAL ILLUSTRATION. Patient With Suspected Myocarditis Who Was Referred for CMR Evaluation and Event-Free Survival Probability Curve.
This is a case of a 20-year-old male without prior cardiac history who was referred for CMR with suspected myocarditis. The patient presented with chest pain, electrocardiogram changes with inferior ST-segment elevations, elevated troponin, and no angiographically significant coronary artery disease. No recent viral or other illness was known and the patient did not take any medication nor was there illegal substance abuse. EMB was not performed. CMR showed linear, epicardial LGE (A) in the anterior, anterolateral, and inferior/inferolateral segments (white arrows). LGE extent was measured using the FWHM quantification method with an ROI 1 placed in the identified affected myocardium (pink arrow/pink contour). LGE extent presented to be 15.6% (B). On T2-weighted imaging there is increased signal in the same segments where LGE was present (C, white arrows). Signal intensity (SI) ratio of myocardium/skeletal muscle (D, ROI 2 with brown arrow/brown contour) indicates edema in the affected segments with SI ratio being ≥2.0. (E) The event-free (MACE) survival probability curve of patients with suspected myocarditis showed that patients with LGE presence had a significantly worse outcome compared to those without LGE. CMR = cardiac magnetic resonance imaging; EMB = endomyocardial biopsy; FWHM = full width half maximum; LGE = late gadolinium enhancement; MACE = major adverse cardiovascular event; ROI = region of interest..
FIGURE 1. Examples of Different LGE Distributions and Patterns Seen in the Cohort.
(A) Patient #1. A short-axis view is displayed with a patchy LGE distribution and an epicardial and midwall pattern, which can also be depicted in the same patient in a 4-chamber view (B). (C) Patient #2. The short-axis image showing LGE presence in anterolateral, inferolateral, and inferior locations, in a linear distribution and a mostly midwall pattern. (D, E) Patient #3. A patchy distribution with midwall pattern located in the septum is shown in a short-axis view, which was confirmed by a 4-chamber view (E). (F) Patient #4. A diffuse LGE distribution, mainly in the midwall. White arrows describe LGE. LGE = late gadolinium enhancement.
EXTRACELLULAR VOLUME MEASUREMENTS AND T2-WEIGHTED IMAGING.
Since 2009, T1 mapping and extracellular volume (ECV) calculation was incorporated in our protocol and was performed as clinically possible in all referred patients (available in 179; 27% of our patients). For each Look-Locker T1-mapping sequence the endo- and epicardial borders of the LV were traced. Using commercially available software (QMASS MR, Medis Medical Imaging Systems, Leiden, the Netherlands), signal intensity versus time curves were generated from regions of interest in the LV and blood pool. The signal intensity versus time curves for each segment and the blood pool were fitted to an analytical expression for the inversion recovery to obtain T1*, and corrected for the effects of radiofrequency pulse during inversion recovery to calculate T1. The reciprocal of T1 (R1 = 1/T1) was used to plot the myocardial R1 against the R1 in the blood pool. Subsequently, the slope of least-squares regression line for R1 in tissue versus R1 in blood (limited data points with R1 in blood <3.5 s–1) was used to estimate the partition coefficient for gadolinium (λGd). This represents an extension of the formula: λGd = (1/T1Myo Post –1/T1Myo Pre)/(1/T1Blood Post–1/T1Blood Pre). λGd was then multiplied by blood plasma fraction (1 minus the hematocrit expressed as a value between 0 and 1), to obtain segmental myocardial ECV. The global myocardial ECV for an individual was calculated by averaging the myocardial segmental ECV values from all the short-axis slices. Hematocrit was derived from routine blood laboratory testing nearest to the CMR examination. An abnormal elevation of ECV was defined by a 2 SD cutoff of ≥35% (16). T2-weighted inversion recovery prepared fast-spin echo sequence was performed using 3 short-axis slices of 12-mm thickness at the base, mid, and apex and a single long-axis slice in a 4-chamber view were performed in 467 (70%) cases (203 missing). Myocardial edema was evaluated by assessing the ratio of the signal intensity in the different myocardial segments compared to the skeletal muscle (musculus pectoralis major or minor) in each segment as described previously (Central Illustration). A ratio of ≥2 was considered abnormal (6).
FOLLOW-UP OF CLINICAL ENDPOINTS.
All subjects were assessed by an interrogation of Social Security Death Index of the United States and a detailed review of all available electronic medical records. When electronic medical records were insufficient, subjects were evaluated by a standardized checklist-based patient questionnaire by mail and/or followed up by conducting a scripted telephone interview based on the same standardized checklist. The retrospective screening and the follow-up was performed between June 2016 and April 2017. A priori – defined primary major adverse cardiac events (MACE) included: 1) all-cause death; 2) heart failure decompensation requiring hospital admission (17);3) heart transplantation; 4) documented sustained ventricular arrhythmia (>30 s); and 5) recurrent acute myocarditis based on elevated myocardial biomarkers (troponin or creatine kinase), absence of CAD during hospitalization and presence of CMR criteria for myocarditis such as T2-weighted imaging of the myocardium to skeletal muscle ratio ≥2 and LGE in epicardial and or mid-myocardium in nonischemic distribution (6). When more than 1 event occurred in a patient subject, the first event was used. Secondary event included all-cause mortality. All study procedures were approved by our Institutional Review Board in accordance with our institutional guidelines. Informed consent was waived by our Institutional Review Board but all patients had the option of refusing follow-up contact by returning a study letter.
STATISTICAL ANALYSIS.
Categorical variables were presented as percentages of the entire cohort or as a percentage of the corresponding group if relevant data were missing. Continuous variables were expressed as mean ± SD or as median values with interquartile range (IQR) depending on normality of distributions. Categorical variables were compared using the Fisher exact test, whereas comparisons for continuous data were performed using a 2-sample Student t test or Wilcoxon rank-sum test, when appropriate. A 2-sided p value of <0.05 was deemed significant. Time to event was measured from the date of CMR study. Univariable and multivariable associations of risk covariates with clinical events were determined by Cox proportional hazards regression. Survival hazard function curves were displayed using Kaplan-Meier. Interobserver and intraobserver variabilities of CMR parameters were performed using intraclass correlation coefficient (ICC). We performed a clinical multivariable analysis including key clinical features. In addition, we built a multivariable model for MACE that contained a parsimonious set of strongest predictive variables using a backward elimination strategy with a level of stay criteria of p < 0.01. All variables in Tables 1 to 4, except laboratory tests and LGE parameters, were considered. Of note, presence of WMAs, not sublocations, was considered for this model. LGE presence was then added to the multivariable model to assess the prognostic value of LGE, incremental to the variables in the multivariable model. Given that treatment-related clinical risk categories have not been defined in patients with suspected myocarditis, we performed computed continuous net reclassification index (NRI) to determine the incremental value of LGE presence above a set of known clinical risk markers (18). The validity of the proportional-hazards assumption was tested by adding a time-dependent interaction variable for each of the covariates in the model. SAS was used for all statistical analysis version 9.4 (SAS Institute Inc., Cary, North Carolina).
TABLE 1.
Baseline Characteristics
| All Patients (N = 670) | LGE Present (n = 294) | LGE Absent (n = 376) | p Value | |
|---|---|---|---|---|
| Baseline | ||||
| Age, yrs | 47.8 ± 16.0 | 49.2 ± 16.4 | 46.8 ± 15.5 | 0.052 |
| Female | 278 (41) | 90 (31) | 188 (50) | <0.001 |
| Body mass index, kg/m2 | 27.8 ± 6.3 | 27.7 ± 5.9 | 27.8 ± 6.7 | 0.893 |
| Acuteness of presentation | ||||
| Acute chest pain syndromes (<2 weeks) | 350 (52) | 169 (57) | 181 (48) | <0.001 |
| Subacute presentation (≥2 weeks) with dyspnea or left ventricular dysfunction | 201 (30) | 91 (31) | 110 (29) | |
| Subacute presentation (≥2 weeks) with ventricular arrhythmias, syncopal spells, or abnormal ECG | 119 (18) | 34 (12) | 85 (23) | |
| Cardiovascular history | ||||
| Hypertension | 181 (27) | 78 (27) | 103 (27) | 0.953 |
| Tobacco | 76 (11) | 38 (13) | 38 (10) | 0.304 |
| Diabetes | 60 (9) | 22 (8) | 38 (10) | 0.495 |
| Dyslipidemia | 138 (21) | 65 (22) | 73 (19) | 0.636 |
| Medications | ||||
| Aspirin | 186 (28) | 94 (33) | 92 (25) | 0.035 |
| ACE inhibitors | 229 (35) | 114 (40) | 115 (31) | 0.019 |
| Beta-blockers | 266 (40) | 142 (49) | 124 (33) | <0.001 |
| Diuretics | 135 (21) | 78 (27) | 57 (16) | <0.001 |
| Statins | 142 (22) | 74 (26) | 68 (18) | 0.020 |
| Insulin | 23 (4) | 7 (2) | 16 (4) | 0.032 |
| ECG | ||||
| Left bundle branch block | 57 (9) | 27 (9) | 30 (8) | 0.579 |
| Right bundle branch block | 43 (6) | 20 (7) | 23 (6) | 0.719 |
| PR duration, ms | 156 (141–176) | 158 (144–178) | 154 (140–174) | 0.087 |
| PR prolongation (≥200 ms) | 36 (5) | 17 (6) | 19 (5) | 0.684 |
| QRS duration, ms | 99.7 ± 23.3 | 100.1 ± 23.9 | 99.3 ± 22.9 | 0.670 |
| QRS prolongation (≥120 ms) | 89 (13) | 40 (14) | 49 (13) | 0.901 |
| QTc duration, ms | 444.5 ± 40.8 | 449.5 ± 42.3 | 440.5 ± 39.2 | 0.008 |
| QTc prolongation (>470 ms female, >450 ms male) | 241 (36) | 116 (39) | 125 (33) | 0.206 |
| Significant Q-wave | 74 (11) | 38 (13) | 36 (10) | 0.248 |
| ST-segment elevation | 32 (5) | 19 (6) | 13 (4) | 0.087 |
| ST-depression | 26 (4) | 13 (4) | 13 (4) | 0.576 |
| T-wave inversion | 170 (25) | 83 (28) | 87 (23) | 0.198 |
| Low voltage | 46 (7) | 30 (10) | 16 (4) | 0.003 |
| Abnormal ECG | 278 (42) | 130 (44) | 148 (39) | 0.206 |
| Laboratory testing | ||||
| Troponin abnormal | 170 (63) | 104 (73) | 66 (52) | <0.001 |
| Troponin peak, ng/ml | 0.08 (0.0–0.46) | 0.14 (0.0–0.6) | 0.02 (0.0–0.3) | 0.002 |
| Creatine kinase abnormal | 70 (40) | 40 (42) | 30 (38) | 0.580 |
| Creatine kinase peak, U/l | 0.99 ± 0.33 | 1.07 ± 0.36 | 0.91 ± 0.27 | <0.001 |
| White blood cell count abnormal | 105 (35) | 61 (39) | 44 (30) | 0.121 |
| White blood cell count (109/l) | 8.3 (6.6–11.4) | 8.7 (6.7–11.4) | 7.9 (6.4–12.0) | 0.317 |
Values are mean ± SD, n (%), or median (interquartile range). The following numbers were available for the laboratory testing: troponin n = 268 (402 missing), creatine kinase n = 170 (500 missing), white blood cell count n = 302 (368 missing). Frequency data were represented as number of cases (percentage of corresponding group, excluding missing data).
ACE = angiotensin-converting enzyme; ECG = electrocardiogram; LGE = late gadolinium enhancement.
TABLE 4.
Univariable Association for MACE and Death: CMR Variables
| MACE |
Death |
|||
|---|---|---|---|---|
| CMR Potential Predictors | HR (95% CI) | p Value | HR (95% CI) | p Value |
| LVEF, % | 0.95 (0.94–0.97) | <0.001 | 0.96 (0.94–0.98) | <0.001 |
| LVEF <40% | 3.91 (2.07–7.38) | <0.001 | 4.0 (2.95–5.44) | <0.001 |
| LVEDVi, ml/m2 | 1.01 (1.00–1.02) | <0.001 | 1.00 (0.99–1.01) | 0.396 |
| LVEDV, ml | 1.01 (1.00–1.01) | <0.001 | 1.00 (1.00–1.01) | 0.629 |
| LVESVi, ml/m2 | 1.01 (1.01–1.02) | <0.001 | 1.01 (1.00–1.02) | 0.015 |
| LVESV, ml | 1.01 (1.00–1.01) | <0.001 | 1.00 (1.00–1.02) | 0.028 |
| LV mass index, g/m2 | 1.01 (1.00–1.03) | 0.021 | 1.01 (0.99–1.03) | 0.178 |
| LV mass, g | 1.01 (1.00–1.01) | 0.010 | 1.00 (0.99–1.01) | 0.520 |
| RVEF, % | 0.95 (0.93–0.96) | <0.001 | 0.95 (0.93–0.98) | <0.001 |
| RVEDVi, ml/m2 | 1.00 (0.99–1.01) | 0.570 | 1.01 (0.99–1.02) | 0.447 |
| RVEDV, ml | 1.00 (1.00–1.01) | 0.199 | 1.01 (0.99–1.01) | 0.805 |
| RVESVi, ml/m2 | 1.02 (1.01–1.03) | <0.001 | 1.02 (1.01–1.04) | 0.003 |
| RVESV, ml | 1.01 (1.01–1.01) | <0.001 | 1.01 (1.00–1.02) | 0.014 |
| Pericardial effusion | 2.31 (1.54–3.45) | <0.001 | 3.36 (1.81–6.26) | <0.001 |
| Pleural effusion | 4.19 (2.71–6.47) | <0.001 | 5.06 (2.65–9.67) | <0.001 |
| Wall motion abnormalities location | ||||
| Wall motion abnormalities at rest | 3.50 (2.26–5.40) | <0.001 | 4.26 (2.08–8.72) | <0.001 |
| Wall motion abnormalities anterior | 4.09 (2.72–6.14) | <0.001 | 4.26 (2.24–8.08) | <0.001 |
| Wall motion abnormalities lateral | 3.69 (2.46–5.52) | <0.001 | 3.41 (1.82–6.39) | <0.001 |
| Wall motion abnormalities inferior | 4.35 (2.87–6.58) | <0.001 | 4.73 (2.44–9.16) | <0.001 |
| Wall motion abnormalities septal | 4.03 (2.65–6.12) | <0.001 | 3.79 (1.98–7.27) | <0.001 |
| LGE presence | 2.22 (1.47–3.35) | <0.001 | 1.99 (1.05–3.75) | 0.034 |
| LGE location | ||||
| LGE anterior | 1.76 (1.03–3.00) | 0.040 | 2.57 (1.22–5.40) | 0.013 |
| LGE lateral | 1.37 (0.90–2.09) | 0.145 | 1.07 (0.53–2.15) | 0.847 |
| LGE inferior | 1.82 (1.18–2.80) | 0.006 | 1.50 (0.75–3.00) | 0.225 |
| LGE septal | 2.55 (1.77–3.83) | <0.001 | 1.78 (0.92–3.41) | 0.084 |
| LGE mass, g | 1.04 (1.02–1.06) | 0.001 | 1.01 (0.96–1.07) | 0.578 |
| LGE percentage, % | 1.05 (1.02–1.09) | 0.002 | 1.03 (0.98–1.09) | 0.262 |
| LGE extent (per 10% increase) | 1.79 (1.25–2.57) | 0.002 | 1.11 (0.52–2.38) | 0.262 |
| LGE distribution | ||||
| Linear | 1.30 (0.85–2.00) | 0.240 | 1.08 (0.54–2.17) | 0.821 |
| Patchy | 2.93 (1.79–4.80) | <0.001 | 2.73 (1.29–5.80) | 0.009 |
| Diffuse | 1.86 (0.45–7.71) | 0.394 | 2.28 (0.30–17.30) | 0.424 |
| LGE pattern | ||||
| Epicardial | 1.26 (0.82–1.95) | 0.291 | 1.6 (0.84–3.07) | 0.156 |
| Midwall | 2.39 (1.54–3.69) | <0.001 | 1.99 (0.99–4.00) | 0.052 |
| T2-weighted ratio mean | 1.12 (0.84–1.48) | 0.436 | 1.08 (0.70–1.67) | 0.737 |
| T2-weighted ratio abnormal (≥2.0) | 2.14 (1.30–3.52) | 0.003 | 2.82 (1.35–5.92) | 0.006 |
| ECV mean (per 10% increase) | 2.09 (1.07–4.08) | 0.031 | 3.93 (1.11–13.86) | 0.034 |
| ECV mean ≥35% | 3.38 (1.43–7.97) | 0.005 | 5.51 (1.01–30.14) | 0.049 |
RESULTS
PATIENT CHARACTERISTICS.
A total of 744 patients were identified. As shown in Figure 2, 59 (7.9%) were excluded based on CMR findings consistent with myocardial infarction (n = 35), cardiac amyloidosis (n = 6), ventricular noncompaction (n = 3), Takotsubo cardiomyopathy (n = 4), constrictive pericarditis (n = 2), cardiac sarcoidosis (n = 2), and Loeffler endocarditis (n = 2), and 1 each for arrhythmogenic cardiomyopathy, hypertrophic cardiomyopathy, pulmonary embolism, cardiac tumor, and severe valve disease. The remaining 670 formed the study cohort and atherosclerotic coronary disease was considered unlikely to be the etiology of the patients’ symptoms based on the initial cardiac work-up, including an absence of significant stenosis on invasive coronary angiography (n = 213, 32%), a negative stress ECG or nuclear/echocardiographic imaging (n = 192, 29%), a negative coronary computed tomographic angiography (n = 4, 0.6%), and a low CAD likelihood by clinical assessment alone (n = 261, 39%). At a median of 4.7 (IQR: 2.3 to 7.3) years, 2 (0.3%) patients were lost to follow-up. Mean age was 48 ± 16 years and 392 (59%) were male. Two hundred sixty (38.8%) were admitted as inpatients. CMR studies were performed using a 3.0-T scanner in 535 (79.9%) and the remaining using a 1.5-T scanner. Overall baseline characteristics and CMR characteristics including LGE parameters are depicted in Tables 1 and 2. The study cohort consisted of 52% who presented with acute chest pain syndromes, 30% with subacute dyspnea and/or LV dysfunction, and 18% with subacute and atypical signs/symptoms. In this cohort, 376 (56%) of the cohort had LGE negative on CMR but 291 of these 376 (77%) presented with either an acute chest pain syndrome (n = 181, 48%) or subacute dyspnea or LV dysfunction (n = 110, 29%). A minority of patients underwent EMB (n = 57, 9%), and their results were in most cases unspecific, which is consistent with other studies (4–6). Of 479 patients who underwent T2-weighted imaging, 125 (27%) had an abnormal T2-weighted ratio (≥2), with 77 (62%) having no matching abnormal LGE. The median number of days of symptoms before CMR was 7 (IQR: 3 to 14 days). In total, 155 (23%) had a recent infection in the past 3 weeks, and it was more prevalent in patients with the LGE presence (n = 83, 28%) compared to those without LGE (n = 72, 19%, p = 0.007). The median delay between ECG and CMR study was 1 day (IQR 0 to 6 days). Regarding ECG findings, LGE presence was not associated with ST-segment abnormalities, left bundle branch block, or presence of Q-waves (p = 0.680, p = 0.580, and p = 0.260, respectively). However, presence of LGE septal pattern was associated with presence of Q-waves (18% vs. 11%; p = 0.030). The other LGE patterns did not show significant association with any ECG findings (19).
FIGURE 2.
Consort Diagram of Patient Enrollment
ARVC = arrhythmogenic right ventricular cardiomyopathy; CAD = coronary artery disease; CMR = cardiac magnetic resonance; LGE = late gadolinium enhancement.
TABLE 2.
CMR Baseline Characteristics
| All Patients (N = 670) | LGE Present (n = 294) | LGE Absent (n = 376) | p Value | |
|---|---|---|---|---|
| LVEF, % | 49.6 ± 15.0 | 44.7 ± 15.7 | 53.4 ± 13.2 | <0.001 |
| LVEF <40% | 200 (29.9) | 113 (38.4) | 87 (23.1) | <0.001 |
| LVEDVi, ml/m2 | 97.6 ± 33.1 | 105.3 ± 37.9 | 91.4 ± 27.4 | <0.001 |
| LVEDV, ml | 189.1 ± 70.0 | 206.6 ± 81.8 | 175.1 ± 55.2 | <0.001 |
| LVESVi, ml/m2 | 52.8 ± 34.4 | 62.6 ± 39.8 | 44.9 ± 27.0 | <0.001 |
| LVESV, ml | 102.2 ± 68.6 | 122.6 ± 81.8 | 85.9 ± 50.3 | <0.001 |
| LV mass index, g/m2 | 60.6 ± 16.8 | 65.0 ± 18.0 | 56.7 ± 14.8 | <0.001 |
| LV mass, g | 118.1 ± 38.6 | 129.3 ± 40.8 | 109 ± 34.4 | <0.001 |
| RVEF, % | 48.8 ± 11.1 | 45.8 ± 12.9 | 51.4 ± 8.6 | <0.001 |
| RVEDVi, ml/m2 | 79.9 ± 21.3 | 82.0 ± 21.8 | 78.1 ± 20.7 | 0.027 |
| RVEDV, ml | 155.6 ± 50.6 | 162.1 ± 53.3 | 150.0 ± 47.6 | 0.004 |
| RVESVi, ml/m2 | 41.7 ± 17.4 | 45.7 ± 20.3 | 38.3 ± 13.8 | <0.001 |
| RVESV, ml | 81.4 ± 38.1 | 90.4 ± 44.5 | 73.9 ± 29.8 | <0.001 |
| Pericardial effusion | 169 (25) | 96 (33) | 73 (19) | <0.001 |
| Pleural effusion | 83 (12) | 48 (16) | 35 (9) | 0.006 |
| Wall motion abnormalities location | ||||
| Wall motion abnormalities at rest | 280 (42) | 181 (62) | 99 (26) | <0.001 |
| Wall motion abnormalities anterior | 203 (30) | 128 (44) | 75 (20) | <0.001 |
| Wall motion abnormalities lateral | 203 (30) | 132 (45) | 71 (19) | <0.001 |
| Wall motion abnormalities inferior | 213 (32) | 135 (46) | 78 (21) | <0.001 |
| Wall motion abnormalities septal | 235 (35) | 145 (49) | 90 (24) | <0.001 |
| LGE location | ||||
| LGE anterior | 70 (10) | 70 (24) | – | – |
| LGE lateral | 177 (26) | 177 (60) | – | – |
| LGE inferior | 137 (20) | 137 (47) | – | – |
| LGE septal | 171 (26) | 171 (58) | – | – |
| LGE mass, g | 2.6 ± 5.4 | 6.2 ± 6.8 | – | – |
| LGE mass, % | 2.2 ± 4.4 | 5.1 ± 5.4 | – | – |
| LGE distribution | – | – | ||
| Linear | 166 (25) | 166 (57) | – | – |
| Patchy | 117 (18) | 117 (40) | – | – |
| Diffuse | 11 (2) | 11 (4) | – | – |
| LGE pattern | – | – | ||
| Epicardial | 168 (25) | 168 (57) | – | – |
| Midwall | 115 (17) | 115 (39) | – | – |
| Transmural (diffuse distribution) | 11 (2) | 11 (4) | – | – |
| T2-weighted ratio mean | 1.80 ± 0.79 | 1.80 ± 0.75 | 1.81 ± 0.82 | 0.857 |
| T2-weighted ratio abnormal (≥2.0) | 125 (27) | 63 (29) | 62 (25) | 0.230 |
| ECV mean (per 10% increase) | 3.2 ± 0.60 | 3.3 ± 0.60 | 3.1 ± 0.60 | 0.004 |
| ECV mean ≥35% | 50 (27) | 28 (35) | 22 (22) | 0.066 |
Values are mean ± SD or n (%). The following data for T2-weighted imaging and ECV were available: T2-weighted n = 467 (203 missing), ECV n = 191 (479 missing). Frequency data were represented as number of cases (percentage of corresponding group, excluding missing data).
CMR = cardiac magnetic resonance; ECV = extracellular volume; LGE = late gadolinium enhancement; LV = left ventricular; LVEDV = left ventricular end diastolic volume; LVEDVi = left ventricular end-diastolic volume index; LVEF = left ventricular ejection fraction; LVESV = left ventricular end-systolic volume; LVESVi = left ventricular end-systolic volume index; RVEDV = right ventricular end-diastolic volume; RVEDVi = right ventricular end-diastolic volume index; RVEF = right ventricular ejection fraction; RVESV = right ventricular end-systolic volume; RVESVi = right ventricular end-systolic volume index.
The interobserver ICCs were 0.88, 0.93, 0.75, and 0.93 for LGE, LV ejection fraction (LVEF), T2-weighted ratio, and ECV, respectively. The intra-observer ICCs were 0.89, 0.96, 0.82, and 0.94 for LGE, LVEF, T2-weighted ratio, and ECV, respectively.
UNIVARIABLE AND MULTIVARIABLE ASSOCIATIONS WITH MACE, DEATH, AND EVENT-FREE SURVIVAL.
MACE occurred in 98 (15%) patients including 29 deaths (4%), 38 (6%) heart failure hospitalizations, 22 (3%) cases of sustained ventricular arrhythmia, 7 (1%) recurrent myocarditis, and 2 (0.3%) heart transplantations. In 294 (44%) patients, LGE was present on CMR. The event-free probability curves in the Central Illustration show that patients with LGE presence were associated with significant increased hazards to MACE (p < 0.001) and death (p = 0.034). Annualized event rates dichotomized by LGE presence were 2.1% and 4.8%, respectively (p < 0.001) for MACE and 0.9% and 1.7% (p = 0.027), respectively, for death (Figure 3). In the event-free MACE probability curve (Figure 4) of patients with LVEF ≥40%, LGE presence maintains its prognostic association with MACE (p = 0.004). However, patients with LVEF <40% had worse outcome compared to patients with LVEF ≥40% (p = 0.008), and the addition of LGE presence did not provide additional prognostic association with MACE in the subgroup of LVEF <40% (Figure 4). Figure 5 shows the annualized event rates stratified by LGE presence and LVEF dichotomized by a 40% cutoff. Subjects with LVEF <40% and LGE presence experienced markedly higher cardiac events. In subjects with LVEF ≥40%, annualized event rates increased from 1.1% in LGE absence to 2.6% in patients with LGE presence (p = 0.004). In subjects with LGE presence, the annualized event rates escalated from 2.6% to 6.4% (p = 0.002) with LVEF $40% and LVEF <40%, respectively. Univariable association of clinical and CMR variables are shown in Tables 3 and 4. In Table 5, univariable association of LGE parameters for MACE based on acuteness of presentation (acute vs. subacute) is presented. With regard to myocardial pattern of LGE, midwall and patchy involvement showed a more than 2-fold increased hazards to MACE. Septal LGE location showed strong association with MACE, whereas lateral location did not show significant association with MACE. When all LGE locations (anterior, inferior, septal, and lateral) were entered into a Cox regression model using a stepwise forward selection strategy using p < 0.05 as criteria for model entry, only septal location of LGE was selected. When all LGE distributions (linear, patchy, and diffuse) and all LGE patterns (epicardial, midwall, and transmural) were entered into a Cox regression model using a stepwise forward selection strategy using p < 0.05 as criteria for model entry, only mid-wall fibrosis was selected. In the multivariable clinical model including age, sex, body mass index (kg/m2), dyspnea, diuretics, LVEF 40%, WMA, QTc, and LGE presence, LGE presence maintained significant association with MACE, with an adjusted hazard ratio (HR) of 1.72 (95% confidence interval [CI]: 1.08 to 2.76; p = 0.023). LGE extent per 10% increase was associated with a 79% increase in risk of MACE. Mean T2-weighted ratio was not associated with MACE or death (HR: 1.12; 95% CI: 0.84 to 1.48; p = 0.436 and HR: 1.08; 95% CI: 0.70 to 1.67; p = 0.737, respectively). However, an abnormal T2-weighted ratio (≥2.0) was significantly associated with outcome (MACE HR: 2.14; 95% CI: 1.30 to 3.52; p = 0.003 and death HR: 2.82; 95% CI: 1.35 to 5.92; p = 0.006). ECV calculation using T1 mapping was performed in a subset of 189 subjects (28%), of which 10 were excluded due to low image quality. ECV mean (per 10% increase) was associated with a >2-fold and near 4-fold increased MACE and death, respectively. A “normal” CMR, defined by a LVEF ≥55%, no WMA, and no LGE, portends to a very low annual event rates of MACE and death, compared to subjects with an “abnormal” CMR (0.9% vs. 4.7%, and 0.4% vs. 1.7% for MACE and death, respectively).
FIGURE 3. Annualized Event Rates Between LGE Presence and LGE Absence in Patients With Suspected Myocarditis.
There are significant differences between the annualized event rates for MACE and death between patients with suspected myocarditis and LGE presence versus those without LGE. LGE = late gadolinium enhancement; MACE = major adverse cardiac event(s).
FIGURE 4. The Event-Free (MACE) Survival Probability Curve of Patients With Suspected Myocarditis and Combination of LGE Presence and LVEF Are Displayed.
Patients with LVEF ≥40% and LGE absence have a significantly better prognosis compared to those with LVEF ≥40% and LGE presence. However, patients with LVEF <40% and LGE absence or presence have the worst outcome compared to those with LVEF ≥40%. LVEF = left ventricular ejection fraction; other abbreviations as in Figures 1 and 3.
FIGURE 5. Annualized Event Rates Between LGE Presence and LGE Absence and LVEF in Patients With Suspected Myocarditis.
There are significant differences between the annualized event rates for MACE and death between patients with suspected myocarditis and LVEF ≥40% with LGE presence versus those without LGE. In patients with suspected myocarditis and LVEF <40%, annualized event rates for MACE and death were not significantly different between patients with LGE present versus patients without LGE. Abbreviations as in Figures 1, 3, and 4..
TABLE 3.
Univariable Association for MACE and Death: Clinical Variables
| MACE |
Death |
|||
|---|---|---|---|---|
| Potential Predictors | HR (95% CI) | p Value | HR (95% CI) | p Value |
| Baseline | ||||
| Age, yrs | 1.03 (1.01–1.04) | <0.001 | 1.06 (1.04–1.08) | <0.001 |
| Female | 1.60 (1.07–2.38) | 0.021 | 1.75 (0.94–3.27) | 0.077 |
| Body mass index, kg/m2 | 1.05 (1.02–1.08) | 0.001 | 1.00 (0.95–1.05) | 0.991 |
| Referral reasons | ||||
| Acute (<2 weeks) vs. subacute presentation (≥2 weeks) | 1.87 (1.22–2.86) | 0.003 | 1.39 (0.73–2.64) | 0.315 |
| History | ||||
| Hypertension | 1.72 (1.14–2.61) | 0.011 | 2.04 (1.05–4.24) | 0.035 |
| Tobacco | 1.59 (0.99–2.58) | 0.057 | 2.11 (0.96–2.00) | 0.079 |
| Diabetes | 2.51 (1.49–4.22) | 0.001 | 1.26 (0.47–3.39) | 0.651 |
| Dyslipidemia | 1.46 (0.93–2.28) | 0.101 | 1.82 (0.94–3.54) | 0.076 |
| Medications | ||||
| Aspirin | 1.47 (1.19–1.82) | <0.001 | 1.35 (1.07–1.72) | 0.012 |
| ACE inhibitors | 1.80 (1.21–2.68) | 0.004 | 1.42 (0.76–2.65) | 0.268 |
| Beta-blockers | 2.34 (1.55–3.51) | <0.001 | 3.70 (1.88–7.28) | <0.001 |
| Diuretics | 3.03 (2.01–4.56) | <0.001 | 3.34 (1.79–6.26) | <0.001 |
| Statins | 1.50 (0.95–2.35) | 0.080 | 1.29 (0.63–2.64) | 0.485 |
| Insulin | 3.62 (1.82–7.21) | <0.001 | 2.5 (0.77–8.12) | 0.128 |
| ECG | ||||
| Left bundle branch block | 0.81 (0.37–1.74) | 0.585 | 0.27 (0.04–1.96) | 0.196 |
| Right bundle branch block | 1.04 (0.48–2.25) | 0.917 | 1.90 (0.74–4.86) | 0.180 |
| PR duration, ms | 1.00 (0.99–1.00) | 0.758 | 0.99 (0.98–1.00) | 0.207 |
| PR prolongation (≥200 ms) | 1.47 (0.68–3.20) | 0.332 | 1.61 (0.49–5.32) | 0.435 |
| QRS duration, ms | 1.00 (0.99–1.01) | 0.758 | 1.00 (0.98–1.01) | 0.649 |
| QRS prolongation (≥120 ms) | 0.90 (0.49–1.66) | 0.746 | 0.76 (0.27–2.16) | 0.609 |
| QTc duration, ms | 1.01 (1.01–1.02) | <0.001 | 1.01 (1.01–1.02) | <0.001 |
| QTc prolongation (>470 ms female, >450 ms male) | 2.25 (1.47–3.45) | <0.001 | 2.77 (1.39–5.52) | 0.004 |
| Significant Q-wave | 1.40 (0.81–2.44) | 0.230 | 1.66 (0.73–3.78) | 0.229 |
| ST-segment elevation | 0.47 (0.15–1.48) | 0.195 | 0.39 (0.05–2.83) | 0.350 |
| ST-depression | 0.61 (0.19–1.94) | 0.404 | 1.04 (0.25–4.35) | 0.953 |
| T-wave inversion | 1.26 (0.82–1.94) | 0.295 | 0.82 (0.40–1.70) | 0.594 |
| Low voltage | 1.92 (1.05–3.52) | 0.034 | 1.39 (0.49–3.91) | 0.532 |
| Abnormal ECG | 1.16 (0.78–1.74) | 0.455 | 0.93 (0.50–1.75) | 0.824 |
| Laboratory tests | ||||
| Troponin abnormal | 1.01 (0.57–1.79) | 0.968 | 1.54 (0.50–4.79) | 0.454 |
| Troponin peak, ng/ml | 1.07 (0.90–1.27) | 0.467 | 0.93 (0.56–1.53) | 0.774 |
| Creatine kinase abnormal | 1.19 (0.62–2.29) | 0.596 | 2.30 (0.66–7.99) | 0.188 |
| Creatine kinase peak, U/l | 1.00 (1.00–1.00) | 0.594 | 1.00 (1.00–1.00) | 0.819 |
| White blood cell count abnormal | 1.79 (1.09–2.92) | 0.021 | 1.11 (0.49–2.50) | 0.810 |
| White blood cell count peak, 109/l | 1.02 (0.98–1.06) | 0.406 | 0.98 (0.89–1.07) | 0.636 |
The following numbers were available for the laboratory testing: troponin n = 268 (402 missing), creatine kinase n = 170 (500 missing), white blood cell count n = 302 (368 missing).
CI = confidence intervals; HR = hazard ratio; MACE = major adverse cardiovascular event; other abbreviations as in Table 1.
TABLE 5.
Univariable Association of LGE Parameters for MACE Based on Acuteness of Presentation
| Acute Presentation (n = 350) |
Subacute Presentation (n = 320) |
|||
|---|---|---|---|---|
| CMR Potential Predictors | HR (95% CI) | p Value | HR (95% CI) | p Value |
| LGE presence | 1.97 (1.20–3.25) | 0.008 | 2.43 (1.18–5.00) | 0.016 |
| LGE percentage | 1.74 (1.18–2.58) | 0.005 | 1.26 (0.45–3.53) | 0.664 |
| LGE location | ||||
| LGE anterior | 1.42 (0.76–2.66) | 0.275 | 2.09 (0.73–5.99) | 0.169 |
| LGE lateral | 1.38 (0.83–2.28) | 0.212 | 1.11 (0.50–2.49) | 0.798 |
| LGE inferior | 1.80 (1.08–3.01) | 0.024 | 1.59 (0.71–3.56) | 0.258 |
| LGE septal | 2.49 (1.53–4.07) | <0.001 | 2.40 (1.16–4.95) | 0.017 |
| LGE distribution | ||||
| Linear | 1.50 (0.90–2.49) | 0.119 | 0.83 (0.36–1.94) | 0.670 |
| Patchy | 1.77 (1.01–3.07) | 0.044 | 3.86 (1.83–8.14) | <0.001 |
| Diffuse | 0.63 (0.09–4.58) | 0.651 | 3.84 (0.52–28.25) | 0.187 |
| LGE pattern | ||||
| Epicardial | 1.23 (0.74–2.06) | 0.426 | 1.07 (0.46–2.48) | 0.877 |
| Midwall | 2.15 (1.24–3.70) | 0.006 | 2.88 (1.38–6.01) | 0.005 |
| Transmural | 2.83 (0.69–11.61) | 0.149 | – | – |
One hundred nineteen patients (18%) were considered low risk given their subacute and atypical symptoms and normal LV function. When these 119 patients were excluded, the robust associations of LGE presence and LGE extent (per 10%) with MACE persisted, and they corresponded to a 2-fold (HR: 2.02; p = 0.001) and a 69% (HR: 1.69; p = 0.005) increase in MACE, respectively. A septal location and a midwall pattern remained the features of highest risks (HR: 2.30 and 2.35; both p = 0.0001).
RISK RECLASSIFICATION AND NET RECLASSIFICATION IMPROVEMENT.
Using a backward elimination regression strategy, a parsimonious set of multivariable predictors formed a clinical model that included patient age in years, body mass index (kg/m2), right ventricular ejection fraction (RVEF) (%), and LVEF <40%. When LGE presence was added, the continuous NRI was 0.39 (95% CI: 0.10 to 0.67) with the proportion of events and nonevents correctly reclassified in 22.8% and 16.6% of the cases, respectively. The validity of the proportional-hazards assumption was tested valid for all variables in this model.
DISCUSSION
Diagnosis of acute myocarditis remains challenging due to a lack of clinical reference standard and the nonspecificities of presenting signs and symptoms and ECG findings. In experienced centers, CMR has become a routine clinical investigation; however, the current prognostic evidence of this approach is limited. Our consecutive cohort represents the largest effort to-date when patients were referred to undergo CMR for suspected myocarditis as the primary concern. We showed that key CMR variables of tissue characterization provide strong prognostic values in risk stratifying patients in this heterogeneous clinical setting.
Other studies had reported prognostic association of LGE but in more narrowed clinical settings. Grün et al. (3) studied 222 pre-selected endomyocardial biopsy-proven myocarditis patients with CMR and they observed that LGE was the strongest independent predictor of all-cause mortality in this high-risk cohort with moderate LV dysfunction. In another study of 58 pediatric patients with myocarditis, LVEF and LGE both showed independent association with outcome (20). However, the pediatric cohort generally has a more fulminant course of myocarditis (21) and such results might not extrapolate to an adult population. Other studies used CMR criteria for myocarditis for study inclusion and thus could not fully represent the diverse setting when CMR was the key noninvasive modality called upon to diagnose and prognosticate patient risk. Chopra et al. (22) assessed 112 patients with a CMR-based diagnosis of myocarditis, and reported that an infarct-like LGE pattern were more likely to show larger LGE extent, lower LVEF, lower RVEF, and a greater risk of MACE at a short-term follow-up of 16 months. Similarly, the study by Sanguineti et al. (23) included 203 patients with myocarditis based on CMR criteria and it showed that LVEF was a predictor for MACE in adjusted analysis, whereas LGE-based variables were not. Our study design targeted the more common practice setting when consecutive patients with suspected myocarditis were referred to CMR for diagnosis and risk stratification, regardless of any decision to perform EMB. In fact, EMB is no longer routinely performed in patients with suspected myocarditis in most centers given its high false negative rate in this clinical setting (4,5).
A typical epicardial or midwall LGE pattern has been associated with the diagnosis of myocarditis (24), but prognosis implications of myocardial patterns have not been reported. We showed that patients having midwall and septal LGE involvement showed a higher risk for MACE. This finding is consistent with the prior report from Mahrholdt et al. (25) who showed that LGE involvement of the septal wall was associated with infection from the more serious human herpesvirus 6 and persistent LV dysfunction, whereas LGE involvement of the lateral wall was associated with the more benign parvovirus B19 and better healing at follow-up. Schumm et al. (26) studied 405 patients with clinically suspected myocarditis and reported favorable prognosis in patients who had a normal CMR. The study by Schumm et al. (26) included patients with infiltrative diseases which take on a different natural history than patients with myocarditis. In comparison to their study, we confirmed the favorable outcomes of a normal CMR study in our larger study cohort of patients with a primary suspicion of myocarditis, but also observed robust prognostic value of LGE incremental to LVEF. We performed LGE quantification and observed that there is a role for quantitative analysis of LGE imaging in patients with suspected myocarditis. Finally, pericardial and pleural effusion detected by CMR, previously discussed as additional diagnostic criteria for myocarditis (27), are significant risk markers probably representing a more serious concurrent serositis.
T2-weighted imaging, a pulse sequence sensitive to regional or global increases of myocardial water (interstitial edema) that are known to be an integral part of the inflammatory response, is part of the Lake Louise criteria and is considered to help establish the diagnosis of myocarditis (6). We observed in our subgroup analysis that there is also an association of abnormal T2-weighted imaging with outcome. Because myocarditis is considered to be a diffuse disease with inflammation of the myocardium, recent trials using the newer approaches of native T1 mapping and ECV were shown to improve the diagnostic accuracy of myocarditis compared to the standard Lake Louise criteria (28). In other patient cohorts, such as sarcoid patients or systemic inflammatory disease patients (e.g., systemic lupus erythematosus) recent studies showed that mapping techniques and ECV were superior to other CMR features in detecting diseased patients (29,30). However, a challenge for the CMR community is to decipher whether such findings in myocarditis or other inflammatory diseases represent active inflammation or chronic fibrosis, or indeed both. In our subset analysis, ECV showed a significant association with MACE and death and might be of additional value for outcome prediction in patients with suspected myocarditis. However, to make definite conclusions, larger studies are needed to support these findings.
STUDY LIMITATIONS.
First, our study has the limitations from a retrospective design without a strategic randomization to any specific therapy. Consequently, potential biases introduced by CMR findings to patient outcomes due to medical or procedural therapies exist. Second, assessing the clinical causes of patient deaths retrospectively is imprecise, so our study was not powered to inform regarding the association of CMR findings with cardiac or arrhythmic deaths. Third, early gadolinium enhancement (EGE) has inconsistent image quality and reproducibility is challenging for technical reasons (timing of image acquisition); therefore, EGE is not routinely used in many centers (31). Furthermore, EGE has lower accuracy than LGE and T2-weighted imaging in diagnosing myocarditis (31). Although EGE is described in the Lake Louise criteria (6), our CMR protocol did not include EGE imaging. Last, T2-weighted images are often affected by artifacts which may introduce inaccuracies to T2-weighted ratio. Further technical development of T2 mapping may be promising in scaling the severity of myocardial edema. Lastly, risk characteristics of our cohort were influenced by local use of CMR imaging in this clinical setting and this may constitute a source of referral bias.
CONCLUSIONS
CMR tissue characterization provides effective risk stratification in patients with suspected myocarditis.
PERSPECTIVES.
COMPETENCY IN MEDICAL KNOWLEDGE:
In patients with suspected myocarditis, tissue characterization by CMR provides effective risk stratification incremental to key clinical markers including LV function. Patients with suspected myocarditis and a normal CMR study have a favorable outcome.
TRANSLATIONAL OUTLOOK:
Future studies may identify genetic markers that predict the course of patients with suspected myocarditis and may help to develop new therapeutic strategies to treat the disease.
Acknowledgments
Dr. Gräni receives funding support from the Novartis Foundation for Medical-Biological Research, Bangerter-Rhyner Foundation, Swiss Sports Medicine Society, and Kreislauf Kardiologie Foundation. Dr. Murthy owns stock in General Electric; receives speaking fees from Bracco Diagnostics; and received research funding from INVIA Medical Imaging Solutions and 1R01HL136685 from NHLBI.
ABBREVIATIONS AND ACRONYMS
- CMR
cardiac magnetic resonance imaging
- CI
confidence interval
- ECG
electrocardiogram
- ECV
extracellular volume
- EMB
endomyocardial biopsy
- HR
hazard ratio
- ICC
intraclass correlation coefficient
- IQR
interquartile range
- LGE
late gadolinium enhancement
- LV
left ventricular
- LVEF
left ventricular ejection fraction
- MACE
major adverse cardiovascular event(s)
- TE
echo time
- TR
repetition time
- WMA
wall motion abnormality
Footnotes
All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
REFERENCES
- 1.Basso C, Calabrese F, Corrado D, Thiene G. Postmortem diagnosis in sudden cardiac death victims: macroscopic, microscopic and molecular findings. Cardiovasc Res 2001;50:290–300. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Cooper LT Jr., Myocarditis. N Engl J Med 2009; 360:1526–38. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Grün S, Schumm J, Greulich S, et al. Long-term follow-up of biopsy-proven viral myocarditis: predictors of mortality and incomplete recovery. J Am Coll Cardiol 2012;59:1604–15. [DOI] [PubMed] [Google Scholar]
- 4.Baughman KL. Diagnosis of myocarditis: death of Dallas criteria. Circulation 2006;113: 593–5. [DOI] [PubMed] [Google Scholar]
- 5.Kindermann I, Kindermann M, Kandolf R, et al. Predictors of outcome in patients with suspected myocarditis. Circulation 2008;118:639–48. [DOI] [PubMed] [Google Scholar]
- 6.Friedrich MG, Sechtem U, Schulz-Menger J, et al. Cardiovascular magnetic resonance in myocarditis: a JACC White Paper. J Am Coll Cardiol 2009;53:1475–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Madias JE. Transient attenuation of the amplitude of the QRS complexes in the diagnosis of Takotsubo syndrome. Eur Heart J Acute Cardiovasc Care 2014;3:28–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Straus SM, Kors JA, De Bruin ML, et al. Pro-longed QTc interval and risk of sudden cardiac death in a population of older adults. J Am Coll Cardiol 2006;47:362–7. [DOI] [PubMed] [Google Scholar]
- 9.Hancock EW, Deal BJ, Mirvis DM, et al. AHA/ ACCF/HRS recommendations for the standardization and interpretation of the electrocardiogram: part V: electrocardiogram changes associated with cardiac chamber hypertrophy: a scientific statement from the American Heart Association Electrocardiography and Arrhythmias Committee, Council on Clinical Cardiology; the American College of Cardiology Foundation; and the Heart Rhythm Society. Endorsed by the International Society for Computerized Electrocardiology. J Am Coll Cardiol 2009;53:992–1002. [DOI] [PubMed] [Google Scholar]
- 10.Konno T, Shimizu M, Ino H, et al. Diagnostic value of abnormal Q waves for identification of preclinical carriers of hypertrophic cardiomyopathy based on a molecular genetic diagnosis. Eur Heart J 2004;25:246–51. [DOI] [PubMed] [Google Scholar]
- 11.De Lazzari M, Zorzi A, Baritussio A, et al. Relationship between T-wave inversion and transmural myocardial edema as evidenced by cardiac magnetic resonance in patients with clinically suspected acute myocarditis: clinical and prognostic implications. J Electrocardiol 2016;49:587–95. [DOI] [PubMed] [Google Scholar]
- 12.Thygesen K, Alpert JS, Jaffe AS, et al. Third universal definition of myocardial infarction. J Am Coll Cardiol 2012;60:1581–98. [DOI] [PubMed] [Google Scholar]
- 13.Kramer CM, Barkhausen J, Flamm SD, Kim RJ, Nagel E. Standardized cardiovascular magnetic resonance (CMR) protocols 2013 update. J Cardiovasc Magn Reson 2013;15:91. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Coelho-Filho OR, Mongeon FP, Mitchell R, et al. Role of transcytolemmal water-exchange in magnetic resonance measurements of diffuse myocardial fibrosis in hypertensive heart disease. Circ Cardiovasc Imaging 2013;6:134–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Flett AS, Hasleton J, Cook C, et al. Evaluation of techniques for the quantification of myocardial scar of differing etiology using cardiac magnetic resonance. J Am Coll Cardiol Img 2011;4:150–6. [DOI] [PubMed] [Google Scholar]
- 16.Neilan TG, Coelho-Filho OR, Shah RV, et al. Myocardial extracellular volume fraction from T1 measurements in healthy volunteers and mice: relationship to aging and cardiac dimensions. J Am Coll Cardiol Img 2013;6:672–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Ponikowski P, Voors AA, Anker SD, et al. 2016 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure: the Task Force for the diagnosis and treatment of acute and chronic heart failure of the European Society of Cardiology (ESC)Developed with the special contribution of the Heart Failure Association (HFA) of the ESC. Eur Heart J 2016;37:2129–200. [DOI] [PubMed] [Google Scholar]
- 18.Pencina MJ, D’Agostino RB Sr., D’Agostino RB Jr., Vasan RS. Evaluating the added predictive ability of a new marker: from area under the ROC curve to reclassification and beyond. Stat Med 2008;27: 157–72; discussion 207–12. [DOI] [PubMed] [Google Scholar]
- 19.Deluigi CC, Ong P, Hill S, et al. ECG findings in comparison to cardiovascular MR imaging in viral myocarditis. Int J Cardiol 2013;165:100–6. [DOI] [PubMed] [Google Scholar]
- 20.Sachdeva S, Song X, Dham N, Heath DM, DeBiasi RL. Analysis of clinical parameters and cardiac magnetic resonance imaging as predictors of outcome in pediatric myocarditis. Am J Cardiol 2015;115:499–504. [DOI] [PubMed] [Google Scholar]
- 21.Amabile N, Fraisse A, Bouvenot J, Chetaille P, Ovaert C. Outcome of acute fulminant myocarditis in children. Heart 2006;92:1269–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Chopra H, Arangalage D, Bouleti C, et al. Prognostic value of the infarct- and non-infarct like patterns and cardiovascular magnetic resonance parameters on long-term outcome of patients after acute myocarditis. Int J Cardiol 2016;212:63–9. [DOI] [PubMed] [Google Scholar]
- 23.Sanguineti F, Garot P, Mana M, et al. Cardio-vascular magnetic resonance predictors of clinical outcome in patients with suspected acute myocarditis. J Cardiovasc Magn Reson 2015;17:78. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Jeserich M, Konstantinides S, Pavlik G, Bode C, Geibel A. Non-invasive imaging in the diagnosis of acute viral myocarditis. Clin Res Cardiol 2009;98: 753–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Mahrholdt H, Wagner A, Deluigi CC, et al. Presentation, patterns of myocardial damage, and clinical course of viral myocarditis. Circulation 2006;114:1581–90. [DOI] [PubMed] [Google Scholar]
- 26.Schumm J, Greulich S, Wagner A, et al. Cardiovascular magnetic resonance risk stratification in patients with clinically suspected myocarditis. J Cardiovasc Magn Reson 2014;16:14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Ong P, Athansiadis A, Hill S, et al. Usefulness of pericardial effusion as new diagnostic criterion for noninvasive detection of myocarditis. Am J Cardiol 2011;108:445–52. [DOI] [PubMed] [Google Scholar]
- 28.Lurz P, Luecke C, Eitel I, et al. Comprehensive cardiac magnetic resonance imaging in patients with suspected myocarditis: the MyoRacer-Trial. J Am Coll Cardiol 2016;67:1800–11. [DOI] [PubMed] [Google Scholar]
- 29.Greulich S, Kitterer D, Latus J, et al. Comprehensive cardiovascular magnetic resonance assessment in patients with sarcoidosis and preserved left ventricular ejection fraction. Circ Cardiovasc Imaging 2016;9 e005022. [DOI] [PubMed] [Google Scholar]
- 30.Puntmann VO, D’Cruz D, Smith Z, et al. Native myocardial T1 mapping by cardiovascular magnetic resonance imaging in subclinical cardiomyopathy in patients with systemic lupus erythematosus. Circ Cardiovasc Imaging 2013;6: 295–301. [DOI] [PubMed] [Google Scholar]
- 31.Luetkens JA, Doerner J, Thomas DK, et al. Acute myocarditis: multiparametric cardiac MR imaging. Radiology 2014;273:383–92. [DOI] [PubMed] [Google Scholar]