Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Molecular Imaging Noninvasive Contrast-Enhanced Ultrasound Molecular Imaging Detects Myocardial Inflammatory Response in Autoimmune Myocarditis David C. Steinl, MSc; Lifen Xu, PhD; Elham Khanicheh, MD, PhD; Elin Ellertsdottir, PhD; Amanda Ochoa-Espinosa, PhD; Martina Mitterhuber, MSc; Katharina Glatz, MD; Gabriela M. Kuster, MD; Beat A. Kaufmann, MD, PhD Downloaded from http://circimaging.ahajournals.org/ by guest on May 12, 2017 Background—Cardiac tests for diagnosing myocarditis lack sensitivity or specificity. We hypothesized that contrastenhanced ultrasound molecular imaging could detect myocardial inflammation and the recruitment of specific cellular subsets of the inflammatory response in murine myocarditis. Methods and Results—Microbubbles (MB) bearing antibodies targeting lymphocyte CD4 (MBCD4), endothelial P-selectin (MBPSel), or isotype control antibody (MBIso) and MB with a negative electric charge for targeting of leukocytes (MBLc) were prepared. Attachment of MBCD4 was validated in vitro using murine spleen CD4+ T cells. Twenty-eight mice were studied after the induction of autoimmune myocarditis by immunization with α-myosin-peptide; 20 mice served as controls. Contrast-enhanced ultrasound molecular imaging of the heart was performed. Left ventricular function was assessed by conventional and deformation echocardiography, and myocarditis severity graded on histology. Animals were grouped into no myocarditis, moderate myocarditis, and severe myocarditis. In vitro, attachment of MBCD4 to CD4+ T cells was significantly greater than of MBIso. Of the left ventricular ejection fraction or strain and strain rate readouts, only longitudinal strain was significantly different from control animals in severe myocarditis. In contrast, contrast-enhanced ultrasound molecular imaging showed increased signals for all targeted MB versus MBIso both in moderate and severe myocarditis, and MBCD4 signal correlated with CD4+ T-lymphocyte infiltration in the myocardium. Conclusions—Contrast-enhanced ultrasound molecular imaging can detect endothelial inflammation and leukocyte infiltration in myocarditis in the absence of a detectable decline in left ventricular performance by functional imaging. In particular, imaging of CD4+ T cells involved in autoimmune responses could be helpful in diagnosing myocarditis. (Circ Cardiovasc Imaging. 2016;9:e004720. DOI: 10.1161/CIRCIMAGING.116.004720.) Key Words: cardiomyopathy ◼ contrast media ◼ echocardiography ◼ leukocyte ◼ myocarditis M imaging methods for the diagnosis and characterization of myocarditis. Echocardiography, radionuclide imaging, and cardiac magnetic resonance imaging have all been used for the detection of myocarditis in clinical and preclinical studies. Echocardiographic features of myocarditis including left ventricular (LV) dilatation, segmental wall motion abnormalities, increased LV sphericity, and transient increases in LV wall thickness7–9 are limited in specificity and hence diagnostic accuracy. For nuclear imaging, Indium-111 radiolabeled antimyosin antibodies have been used for the detection of yocarditis is characterized by an inflammatory infiltration of the myocardium with myocyte degeneration and necrosis in the absence of ischemia.1 Infectious agents, systemic diseases, drugs, and toxins cause myocarditis.2 Myocarditis has been found in 1% to 9% of autopsies and in a higher proportion (5% to 12%) in young individuals with sudden unexplained cardiac death.3,4 This is in contrast to the perceived low frequency of a diagnosis of myocarditis in patients. The clinical presentation of myocarditis is variable, and routine cardiac tests either lack sensitivity (ECG) or specificity (troponins) for diagnosing myocarditis. Endomyocardial biopsy is the gold standard for securing the diagnosis of a suspected myocarditis.5 However, endomyocardial biopsy is invasive, susceptible to sampling error and suffers from low sensitivity and a high interobserver variability in interpretation.6 Thus, there is a need for noninvasive See Editorial by Porter See Clinical Perspective myocardial necrosis in murine models of myocarditis and in clinical studies.10,11 Whereas Indium-111 antimyosin antibody Received February 25, 2016; accepted June 7, 2016. From the Department of Biomedicine (D.C.S., L.X., E.K., E.E., A.O.-E., M.M., G.M.K., B.A.K.), Institute for Pathology University Hospital (K.G.), and Division of Cardiology, University Hospital (G.M.K., B.A.K.), University of Basel, Switzerland. The Data Supplement is available at http://circimaging.ahajournals.org/lookup/suppl/doi:10.1161/CIRCIMAGING.116.004720/-/DC1. Correspondence to Beat A. Kaufmann, MD, PhD, Division of Cardiology, University Hospital Basel, Petersgraben 4, 4031 Basel, Switzerland. E-mail [email protected] © 2016 American Heart Association, Inc. Circ Cardiovasc Imaging is available at http://circimaging.ahajournals.org 1 DOI: 10.1161/CIRCIMAGING.116.004720 2 Steinl et al Ultrasound Molecular Imaging of Myocarditis Downloaded from http://circimaging.ahajournals.org/ by guest on May 12, 2017 imaging has a high sensitivity and negative predictive value, its clinical use is hampered by limited tracer availability, radiation exposure, and delayed imaging to prevent blood pool signal.11 Cardiac magnetic resonance imaging allows detection of myocardial edema or irreversible tissue damage inflicted by the inflammatory infiltrate,12,13 but sensitivity is variable and may be reduced in the absence of tissue necrosis.14 Animal models of both virus-induced and autoimmune myocarditis suggest that heart-specific autoaggressive inflammatory responses are important in the development of chronic cardiomyopathy, and that CD4+ T cells are instrumental in driving the transition to cardiomyopathy.15–17 Recently, fluorine-19 cardiac magnetic resonance has been used for detecting the inflammatory infiltrate in murine autoimmune myocarditis.18 Thus, with the exception of fluorine-19 cardiac magnetic resonance, a common weakness of the aforementioned techniques is that they detect tissue injury (edema, necrosis), which may be absent in lymphocytic myocarditis or which can also be because of, for example, myocardial ischemia, rather than the inflammatory infiltrate. In the past years, microbubble (MB) ultrasound contrast agents have been modified for the purpose of molecular imaging. Either modifications in the shell properties or the conjugation of specific ligands to the MB shell surface have been used to promote adhesion of MB to diseasespecific intravascular events. The feasibility of this strategy for molecular imaging in the myocardium has been proven in vivo.19,20 In this study, we hypothesized that contrast-enhanced ultrasound (CEU) molecular imaging can be used to detect (1) the inflammatory cell recruitment and endothelial inflammatory activation and (2) the recruitment of CD4+ T cells in a murine model of myocarditis. Methods Study Design and Animal Model All experiments were performed in accordance with Swiss Legislation and were approved by the Animal Care Committee of Basel. A total of 52 female BALB/c mice were used. In 4 mice, specific targeting of CD4 with targeted MB was validated in vitro. In 34 mice, experimental autoimmune myocarditis (EAM) was induced by intraperitoneal injection of 400-ng pertussistoxin (List Biological Laboratories) followed 1 hour later by subcutaneous injection of 120 µg of α-myosin heavy chain peptide (α-myosin heavy chain, Ac-RSLKLMATLFSTYASADR-OH, GeneCust) emulsified with complete freund adjuvant (Sigma, 1 mg/mL) on day 0. On day 7, a repeat injection of α-myosin heavy chain only was given. Seven mice injected with saline/complete freund adjuvant served as controls. An additional 7 mice were studied after induction of myocarditis to verify specific targeting of CD4 using blocking antibodies. The immunization with α-myosin heavy chain results in the development of EAM in 60% of injected animals with inflammation peaking around day 21.21 Assessments at that time point included high-resolution echocardiography; strain imaging; measurement of myocardial perfusion; and CEU molecular imaging for leukocytes, P-selectin, and CD4+ lymphocytes. Myocarditis severity was assessed histologically, CD4+ lymphocytes were quantified with immunohistology. MB Preparation Perfluorocarbon-filled, lipid-shelled MB were prepared by sonification of a gas-saturated aqueous suspension of distearoyl phosphatidylcholine (2 mg/mL; Avanti Polar Lipids, Alabaster, AL) and polyoxyethylene-(40)-stearate (1 mg/mL; Sigma). These nontargeted MB were used for the assessment of myocardial perfusion. MB for targeting of leukocytes (MBLc) were prepared by adding distearoylphosphatidylserine (0.3 mg/mL; Avanti Polar Lipids) to the MB shell.22 For targeting of P-selectin and CD4, distearoyl-phosphatidylethanolamine-PEG(3400)-biotin (0.14 mg/mL; Creative PEG Works) was added. MB targeted to P-selectin (MBPSel) and to CD4 (MBCD4) were then prepared by conjugation of biotinylated anti–P-selectin antibody (RB40.34) or anti-CD4 antibody (H129.19) to the MB surface using biotin–streptavidin–biotin linking as previously described.23 Control MB (MBIso) bearing a nonspecific isotype control antibody (R3-34) were also prepared. For in vitro experiments, MB were labeled by incorporation of dioctadecyloxacarbocyanine-perchlorate (DiO, Molecular Probes Inc) into the shell. MB concentration and size were measured by electrozone sensing (Multisizer III, BeckmanCoulter). MB mean (±SD) size was not statistically different for the 4 MB (2.7±0.2 μm for MBIso, 2.5±0.2 μm for MBLc, 2.7±0.2 μm for MBPSel, and 2.7±0.2 μm for MBCD4; P=ns). In Vitro Assessment of MB–CD4 Interaction Flow-chamber experiments were used to assess the specific attachment of MBCD4 to CD4 protein under flow conditions. CD4 protein was adsorbed to culture dishes (35 mm, Corning) at a concentration of 5 ng/µL by overnight incubation. The dishes were then washed with PBS containing 0.05% Tween 20 and blocked with 1% albumin in PBS/0.05% Tween 20. The culture plates were mounted on an inverted parallel plate flow chamber (Glycotech). The flow chamber was placed in an inverted position on a microscope (Olympus BX51W). A suspension of either MBIso or MBCD4 at a concentration of 3×106 mL−1 was drawn through the flow chamber at a shear rate of 2 dynes/cm2. After 5 minutes of continuous flow, the number of attached MB was counted in 20 optical fields (×20 magnification, total area 2.94 mm2). Experiments were performed in triplicate. Specific attachment of MBCD4 to CD4+ lymphocytes was assessed on cell-based adhesion assays. Single-cell suspensions of murine spleen cells from 4 mice were generated using a cell strainer. CD4+ T cells were separated from the cell suspension by depletion of magnetically labeled nontarget cells using a cocktail of biotin-conjugated monoclonal antibodies and antibiotin monoclonal antibodies conjugated to MicroBeads (CD4+ T Cell Isolation Kit II, Miltenyi Biotec). Both the separated nontarget cells and the CD4+ cells were resuspended at 2×107 cells per 100 µL. 1×106 leukocytes (CD4+ or CD4−) were then incubated with 1×107 MB (MBIso or MBCD4) for 10 minutes followed by washing with PBS. Microscopy at ×20 magnification (Zeiss LSM 710) was used to assess MB attachment to the cells with the investigator (E.E.) blinded to all image information. The results are presented as the average number of MB attached per cell. Animal Instrumentation On day 21 after immunization, the mice were anesthetized with 5% isoflurane, which was subsequently reduced to 2% and adapted to stabilize the heart rate between 350 and 450 bpm. The animal core temperature was maintained at 37°C. The chest and neck were depilated. For MB injections, the right internal jugular vein was cannulated and the animals were transferred onto a temperature controlled imaging stage (Vevo Imaging Station). Echocardiography High-frequency (40 MHz, MS 550D transducer) ultrasound imaging (Vevo 2100, VisualSonics Inc) was performed for the assessment of cardiac structure and function. M-Mode images of the LV at the height of the papillary muscles were used to measure anterior and posterior wall thickness, LV end-diastolic diameter, and LV end-systolic diameter. LV ejection fraction and LV mass were derived from these measurements.24 E-wave and A-wave velocities were measured using pulsed Doppler from a modified apical 4-chamber view. Parasternal long- and short-axis B-mode images were acquired at a frame rate of >200 frames per second. Longitudinal, radial, and circumferential global peak strain values were derived from the B-mode images using a speckle-tracking algorithm (VevoStrain, VisualSonics).25 Cine-loops with 2 cardiac cycles not distorted by 3 Steinl et al Ultrasound Molecular Imaging of Myocarditis Downloaded from http://circimaging.ahajournals.org/ by guest on May 12, 2017 Contrast ultrasound imaging (Sequoia Acuson C512; Siemens Medical Systems) was performed with a high-frequency lineararray probe (15L8) held in place by a railed gantry system. The LV was imaged in short axis at the papillary muscle level. ECGgated triggering was used to acquire end-systolic frames. For the detection of the contrast agent, power modulation and pulse inversion imaging at a frequency of 7 MHz and a dynamic range of 50 dB was used. The gain settings were adjusted to levels just below visible speckle and held constant. MBIso, MBLc, MBPSel, and MBCD4 (5×106 MB per injection) were injected in each mouse in random order. Ultrasound imaging was paused from injection until 8 minutes later when imaging was resumed at a mechanical index of 0.87. The first acquired image frame was used to derive the total amount of MB present within the LV myocardium. The MB in the ultrasound beam were then destroyed with several (>10) image frames. Several image frames at a long pulsing interval (every 9 heartbeats) were subsequently acquired to measure signal attributable to freely circulating MB. After log-linear conversion using known dynamic range lookup tables, frames representing freely circulating MB were digitally subtracted from the first image to derive signal from attached MB. Contrast intensity was measured from a region of interest encompassing the whole LV myocardium. The selection of the region of interest was guided by fundamental frequency anatomic images of the LV acquired at 14 MHz. In 7 additional mice with myocarditis, imaging was performed 10 minutes after blocking of CD4 by intravenous injection of antimouse CD4 mAb (H129.19, BD Bioscience; 5 µg). After CEU molecular imaging, microvascular perfusion was assessed using constant infusions of nontargeted MB at an infusion rate of 3×106 per minute. After a high-energy destruction impulse (5 frames, mechanical index of 1.9), end-systolic short-axis image frames of the LV were acquired over several heartbeats (>10) at a low mechanical index of 0.2. On the resulting destruction–replenishment sequences, the average signal intensity of a region of interest placed over the septal, anterior and lateral wall was fitted to the monoexponential function y=A(1−e−ßt), where y is the signal intensity at a given time point, ß is the rate of increase of the curve, which reflects myocardial blood flow velocity, and A is the plateau intensity, which is proportional to myocardial blood volume. Myocardial blood flow was estimated by the product of A*ß. For each animal, 2 to 3 destruction–replenishment curves were averaged.26 CEU imaging data were analyzed blinded to all other experimental data (D.S.). Histology and Immunohistology For the assessment of myocarditis severity, hematoxylin and eosin– stained short-axis cross-sections of the hearts were analyzed by a pathologist blinded for all other experimental data. The cross section of the LV was divided into 4 quadrants by using (1) the papillary muscles and (2) an ink mark on the posterior interventricular sulcus as landmarks. Each of the 4 quadrants was then individually scored on a scale from 0 to 4 with 0 representing no inflammatory infiltrates, 1 small foci of leukocytes between cells, 2 larger foci of >100 leukocytes, 3>10% of the quadrant cross section involved, and 4>30% of the quadrant cross section involved. In addition, an overall score for the entire cross section using the same criteria was given and used for the classification of animals.17 For immunohistochemistry, snap-frozen heart sections were stained with antibodies to CD4 (MCA4635T, AbD Serotec). A fluorescent secondary Statistical Analysis Data were analyzed on GraphPad Prism (version 7). Data are expressed as medians and 25th–75th percentile unless stated otherwise. Comparisons between MB agents in the flow-chamber experiments were made using a Mann–Whitney rank sum test. Attachment of MB to CD4+ versus CD4− cells and MB signals within and between animal groups were compared using a Kruskal–Wallis ANOVA with Dunn post hoc test. Correlations between MB signals and CD4+ cells in tissue were assessed using Spearman correlation. A P value of <0.05 (2 sided) was considered statistically significant. Results In Vitro Assessment of MB–CD4 Interaction The specific targeting of MBLc and MBPSel has been validated previously, whereas this has not been the case for MBCD4.23,27 Therefore, we performed flow chamber and in vitro cell-based assays to characterize MBCD4 interactions with CD4. In a parallel plate flow-chamber system coated with CD4 protein, attachment of MBCD4 was significantly greater than for MBIso at a shear stress encountered in the myocardial microcirculation (Figure 1A). In cell-based assays, MBCD4 attached to CD4+ lymphocytes. Attachment to CD4− cells was 6-fold less, and similar to the attachment of MBIso to both CD4+ and CD4− cells (Figure 1B–1D). Baseline Characteristics and Echocardiographic Parameters Myocarditis developed in 21 of the injected mice (n=7 severity degree 1, n=7 severity degree 2, n=1 severity degree 3, and n=6 A Attached MB / visual field CEU Molecular Imaging and Assessment of Myocardial Perfusion antibody (A-11006 Alexa 488, Invitrogen) was used to visualize CD4, and the cells/mm2 of tissue staining positive for CD4 were quantified. B Attached MB / cell respiratory movements were selected, and the endocardial and epicardial borders were traced on end-diastolic frames. After automated tracing, manual adjustments were made as necessary for good tracking throughout the cardiac cycles. The strain measurements were then averaged over the 2 cardiac cycles without the use of smoothing filters. All echocardiographic images were acquired and analyzed by the same investigator (L.X.) who was blinded to all other experimental data. C MBIso MBCD4 D MBIso MBCD4 CD4-cells CD4+ cells Figure 1. A, Number of microbubbles targeted to CD4 (MBCD4) vs control microbubbles (MBIso) attaching to CD4 protein in a parallel plate flow chamber. Data depict median value (horizontal line), 25% to 75% percentiles (box), and range of values (whiskers). *P<0.0001 vs MBIso. B, Number of MBCD4 and MBIso attaching to murine spleen cells not expressing CD4 vs cells expressing CD4, *P<0.0001 vs all others. Examples showing no attachment of MBCD4 to cells not expressing CD4 (C) and attachment of several MBCD4 per individual CD4+ cell (D), red color indicates cell nuclei and green color indicates microbubbles. MB indicates microbubbles; and MBCD4, MB bearing antibodies targeting lymphocyte CD4. 4 Steinl et al Ultrasound Molecular Imaging of Myocarditis Downloaded from http://circimaging.ahajournals.org/ by guest on May 12, 2017 severity degree 4; Table I in the Data Supplement), whereas n=13 of the injected mice did not develop myocarditis (Figure 2). For all subsequent analyses, mice with severity degrees 3 and 4 were grouped together as severe myocarditis, severity degrees 1 and 2 as moderate myocarditis, whereas injected mice that did not develop myocarditis were grouped together with sham-injected mice as no myocarditis. In animals with myocarditis, the inflammatory infiltrate consisted primarily of mononuclear cells, including macrophages and lymphocytes. Mice with severe myocarditis had lower body weight than both the mice with no and moderate myocarditis (Table). Both the heart weight determined postmortem and the LV mass determined by echocardiography were higher in mice with severe myocarditis compared with the 2 other groups. E-wave velocity was lower in mice with severe myocarditis compared with the other groups. For all other echocardiographic parameters, there were no significant differences between the groups with a trend toward reduced values in animals with severe myocarditis. Of note, both β as an estimate of myocardial blood flow velocity and A*β as an estimate of myocardial blood flow were not significantly different between the groups, albeit with large differences between individual animals. LV Strain Parameters Speckle-tracking–based strain measurements have been reported to be more sensitive for the detection of global changes in LV function when compared with conventional echocardiography.25 We, therefore, assessed global peak strain and peak strain rate on LV long- and short-axis images. In mice with severe myocarditis, radial and circumferential strain showed a trend toward lower values, and longitudinal strain was significantly reduced. However, in mice with moderate myocarditis, strain and strain rate values were identical to the ones in mice without myocarditis (Figure 3). Ultrasound Molecular Imaging Figure 4 shows data for ultrasound molecular imaging in all the 3 groups of mice, and Figure 5 represents examples of colorcoded images for all 4 MB in an animal without, with moderate and with severe myocarditis. Signal enhancement was similarly low for all MB in mice without myocarditis. In contrast, in mice with myocarditis, the signals from MBCD4, MBPSel, and MBLc were all significantly higher than the signal from MBIso. Importantly, this was not just the case in mice with severe myocarditis but also in mice with moderate myocarditis, as well as when comparing mice without myocarditis to mice with myocarditis (Figure 1 in the Data Supplement). In mice with severe myocarditis, an increase in signal from MBIso could be observed, which was, however, not significant when compared with the corresponding signals in the 2 other groups. Whereas the absolute signal intensities for MBCD4, MBPSel, and MBLc increased in mice with severe myocarditis compared with mice with moderate myocarditis, the ratios for the targeted versus the control MB remained constant at a 2- to 3-fold increase because of nonsignificantly elevated signal from MBIso. When examining the performance of ultrasound molecular imaging for diagnosing myocarditis using receiver-operating-characteristic curves, all 3 targeted MB performed better than ejection fraction, longitudinal strain, and MBIso (Figure 2 in the Data Supplement). In addition to in vitro validation, we performed in vivo experiments to confirm the specific attachment of MBCD4 to CD4+ T cells. After injection of a blocking antibody in a dosage sufficient to saturate CD4 on CD4+ T cells in circulation, signal for MBCD4 was abolished in mice with myocarditis (Figure 6). Figure 2. Hematoxylin and eosin–stained histological sections of the myocardium in a control mouse (A and D), a mouse with moderate myocarditis (B and E), and a mouse with severe myocarditis (C and F). The dashed lines (B) denote 2 inflammatory foci that covered ≈8% of the left ventricular area, and the arrowhead (E) denotes a focus of >100 inflammatory cells, (F) shows extensive inflammatory infiltration in a mouse with severe myocarditis. 5 Steinl et al Ultrasound Molecular Imaging of Myocarditis Table. Physiological and Echocardiographic Data No Myocarditis (n=20) Moderate Myocarditis (n=14) Severe Myocarditis (n=7) Body weight (g) 22.5 (18.6–24.8) 20.0 (19.4–22.7) 15.5 (14.9–16.4)* Heart rate (bpm) 388 (349–490) 413 (392–473) 403 (351–461) Heart weight (mg) 112 (105–126) 119 (113–132) 171 (147–207)† 76 (65–90) 88 (72–96) 117 (82–137)‡ LV mass (mg) LV Vol d (µL) 52 (40–57) 46 (44–57) 38 (24–52) LV Vol s (µL) 15 (7–22) 13 (10–18) 9 (7–27) 69 (42–76) EF (%) 73 (61–83) 70 (69–77) −1 E-wave velocity (mm×s ) 613 (563–734) 695 (567–802) 527 (405–561)§ A wave velocity (mm×s ) 415 (371–454) 400 (346–531) 354 (154–428) E/A 1.5 (1.3–1.7) 1.7 (1.3–1.8) 1.5 (1.0–2.0) β (s ) 0.8 (0.5–1.2) 0.6 (0.4–1.0) 0.7 (0.2–0.8) A*β 10 (2–29) 18 (8–31) 26 (6–49) −1 −1 CEU Molecular Imaging Detects Myocarditis Independent of Functional Impairment Despite advances in noninvasive imaging, myocarditis continues to represent a challenging diagnosis. Conventional echocardiography can detect structural and functional alterations in fulminant myocarditis,28 but it is limited in sensitivity and specificity in lower-grade acute myocarditis, which has a higher likelihood of progression to dilated cardiomyopathy.29 In our study, LV mass was increased and LV function was tended to be decreased in the group with severe myocarditis. However, despite the use of high-resolution ultrasound and 80 Strain [%] 25 15 40 10 20 5 0 -40 -20 -30 -15 -20 -10 -10 -5 0 -25 -15 -20 -15 -10 -5 0 Strain rate [s-1] 20 60 radial Discussion Myocarditis and the ensuing inflammatory autoimmune phenomena can lead to chronic dilated cardiomyopathy and heart failure. Currently available noninvasive imaging methods detect the functional or structural consequences of infection or inflammation, such as tissue necrosis or edema, and are limited in diagnostic accuracy. We, therefore, investigated whether CEU molecular imaging can detect the inflammatory process in myocarditis, and whether the specific detection of cellular subsets of the inflammatory infiltrate is possible. Our results indicate that (1) CEU ultrasound molecular imaging detects myocardial inflammatory infiltration and endothelial cell activation independent of functional consequences in autoimmune myocarditis and (2) we show for the first time that CEU molecular imaging allows for the specific detection of CD4+ T cells, which contribute to the pathogenesis of myocarditis and dilated cardiomyopathy. speckle-tracking analysis, there were no differences in LV size and function between the group without and the group with moderate myocarditis. Direct assessment of the inflammatory process responsible not only for acute myocarditis but also for the progression to dilated cardiomyopathy may improve the diagnostic performance of cardiac ultrasound. Our study shows that CEU molecular imaging with MB targeted to leukocytes and to P-selectin yields increased signal not only in the group with severe myocarditis but also in the group with moderate myocarditis, suggesting that the direct detection of the inflammatory infiltrate, rather than its consequences may result in better sensitivity. In addition to relatively nonspecific targeting of activated leukocytes using negatively charged 0 circumferential Immunohistochemistry was used to quantify CD4+ T cells in myocardial tissue to assess whether signal from MBCD4 tracks with the degree of CD4+ T cell infiltration. MBCD4 signal correlated, albeit modestly, with CD4+ cells within the myocardium on histology (Figure 7), this was not the case for MBIso in the same animals. longitudinal Downloaded from http://circimaging.ahajournals.org/ by guest on May 12, 2017 Values are median (25th–75th percentile). A-wave velocity indicates late diastolic mitral inflow velocity during atrial contraction; EF, ejection fraction; E-wave velocity, early diastolic mitral inflow velocity; LV Vol d, left ventricular volume at end diastole; and LV Vol s, left ventricular volume at end systole. *P=0.0014, †P=0.0008, ‡P=0.0134, §P=0.0352 vs no myocarditis. 0 * no Myocarditis n=20 moderate Myocarditis n=14 severe Myocarditis n=7 -10 -5 0 Figure 3. Strain imaging data according to myocarditis severity. Median values (horizontal line), 25% to 75% percentiles (box) and range of values (whiskers) for peak strain (left) and peak strain rate (right) are given. *P=0.0403 vs no myocarditis. 6 Steinl et al Ultrasound Molecular Imaging of Myocarditis P-selectin–targeted MB was greater in severe compared with moderate myocarditis indicates that various degrees of disease activity can be discerned. With regard to the clinical translation of this method for detecting myocarditis, leukocyte-targeted MB perform as well as MB targeted to P-selectin with the advantage of a greater ease of preparation of the contrast agent. CEU Molecular Imaging Assesses the Infiltration With Pathogenic CD4+ Lymphocytes MB,27 we used antibody targeting of the endothelial cell adhesion molecule P-selectin. P-selectin is expressed on the surface of the vascular endothelium on inflammatory stimulation. In myocarditis, P-selectin plays an important role by mediating the recruitment of pathogenic T cells into myocardial tissue.30 Our results are in line with earlier reports that show that CEU molecular imaging can detect cardiac transplant rejection by targeting activated leukocytes or intercellular adhesion molecule-1.31,32 However, whereas the inflammatory process in transplant rejection shares similarities with myocarditis, an important difference that increases the diagnostic challenges is the low grade and focal nature of the inflammatory infiltrate in moderate myocarditis (Figure 2B and 2E). Also, the finding that CEU molecular signal for both leukocyte-targeted and mild myocarditis no myocarditis MBIso severe myocarditis Downloaded from http://circimaging.ahajournals.org/ by guest on May 12, 2017 Figure 4. Signal enhancement from targeted and isotype control microbubbles. Data are median values (horizontal line), 25% to 75% percentiles (box) and range of values (whiskers) of the video intensity. *P=0.027, ‡P=0.0133, †P=0.0004, ¶P=0.0381, ¥P=0.0192, §P=0.0073 vs MBIso in the same animal group. MB indicates microbubbles; MBCD4, MB bearing antibodies targeting lymphocyte CD4; MBPSel, MB bearing endothelial P-selectin; MBIso, MB bearing isotype control antibody; and MBLc, MB with a negative electric charge for targeting of leukocytes. MBCD4 MBPSel In animal models of both viral and autoimmune myocarditis, CD4+ T cells represent a relatively small subset of leukocytes infiltrating the myocardium.33 However, development of myocarditis on transfer of appropriately activated CD4+ T cells in host animals suggests a decisive role of CD4+ T cells.17 In addition, the number of CD4+ T cells within tissue correlates with cardiac functional impairment in the chronic stage of EAM.34 Thus, the specific detection of CD4+ T cells could not only be used to characterize the inflammatory infiltrate in myocarditis but could also be of value in predicting late functional consequences of myocarditis. We, therefore, developed and validated a novel, CD4+-targeted MB using a monoclonal antibody. Flow-chamber assays and in vitro experiments showed that this contrast agent attaches specifically to CD4+ T cells, and that attachment occurs also under flow. In vivo, CEU molecular imaging showed increased signals both in moderate and in severe myocarditis, and the signal correlated with CD4+ T cells present in tissue. Abolishment of signal by blocking of CD4+ (Figure 6) indicates selective targeting similar to in vitro observations. A previous study used imaging of CD8+ T-cell granzyme B activity to assess immune-mediated myocarditis triggered by injection of cytotoxic T cells with fluorescence reflectance imaging.35 However, whereas fluorescence imaging is limited to small-animal models, CEU molecular imaging as used in our study does not have limited penetration and, therefore, could also be feasible in humans. In addition, molecular imaging of specific cellular subsets of MBLC Figure 5. Examples of color-coded contrastenhanced ultrasound (CEU) molecular images in 1 animal without myocarditis (top row), in moderate myocarditis, and in severe myocarditis (bottom row) for microbubble bearing isotype control antibody (MBIso), MB bearing antibodies targeting lymphocyte CD4 (MBCD4), MB bearing endothelial P-selectin (MBPsel), and MB with a negative electric charge for targeting of leukocytes (MBLc). The color scale for the CEU images is shown at the bottom of each frame. 7 Steinl et al Ultrasound Molecular Imaging of Myocarditis Acoustic intensity (AU) 10 8 6 MBIso 4 MBCD4 + H129.19 2 MBLc 0 moderate and severe myocarditis n=7 A Acoustic intensity (AU) the inflammatory immune response in myocarditis may help further elucidating the mechanisms of immune-mediated myocardial disease and be of use in the assessment of treatment effects in the future. There are several limitations of this study that deserve attention. We studied a model of EAM, rather than virusinduced disease. However, EAM shares many pathophysiological features with virus-induced myocarditis, and it has been used extensively to study the autoimmune processes caused by initial viral myocyte infection.36 Regarding CEU molecular imaging, differences in myocardial perfusion between the groups could potentially influence the results. Although there was a high variability in myocardial perfusion in mice with severe myocarditis, and localized perfusion abnormalities in areas with tissue necrosis cannot be excluded in this animal group, we did not find significant 15 B C rs = 0.70 p < 0.001 10 5 0 0 Acoustic intensity (AU) Downloaded from http://circimaging.ahajournals.org/ by guest on May 12, 2017 Figure 6. Blocking of CD4 abolishes signal from microbubble bearing antibodies targeting lymphocyte CD4 (MBCD4). Data are median values (horizontal line), 25% to 75% percentiles (box) and range of values (whiskers) of the video intensity. *P=0.0209 versus MB bearing isotype control antibody (MBIso). MBLc indicates MB with a negative electric charge for targeting of leukocytes. 15 50 100 150 200 CD4+ cells/mm2 50 m 250 D rs = 0.21 p = 0.33 10 5 0 0 50 100 150 200 CD4+ cells/mm2 250 50 m Figure 7. Correlation of contrast-enhanced ultrasound molecular imaging data for MBCD4 (A) and MBIso (B). Examples of extensive myocardial CD4+ T-lymphocyte infiltration (C) and low-level infiltration (D) on immunohistochemistry, green color indicates positive staining for CD4. differences between the groups, and perfusion was identical in mice without and mice with moderate myocarditis. In addition, the low signals for control MB in both the groups with myocarditis argue against an influence of myocardial flow on CEU molecular imaging results. Also, although moderate myocarditis could be detected independent of functional defects, ultrasound molecular imaging may not be sensitive enough to diagnose the early stages of myocarditis when only a few inflammatory foci or minimal myocyte necrosis are present. Finally, the limited spatial resolution of CEU at 7 MHz precluded a regional assessment of disease activity, which would be desirable for the guidance of endomyocardial biopsies in patients. In summary, the results of our study indicate that CEU molecular imaging can be used to detect the inflammatory process in myocarditis and discriminate disease severity independent of cardiac function. Importantly, specific molecular imaging of a leukocyte subset that is driving the autoimmune process in myocarditis is possible. As a consequence, CEU molecular imaging can potentially be useful in diagnosing myocarditis. Future studies will have to address whether imaging of the acute inflammatory response and its components also allows a risk assessment about the future development of dilated cardiomyopathy. Sources of Funding This work was supported by grants from the Swiss National Science Foundation (SNF 310030_149718 to Dr Kaufmann and SNF 310030_144208 to Dr Kuster), the Swiss Life Jubiläumsstiftung and the Kardiovaskuläre Stiftung Basel. Disclosures None. References 1. Caforio AL, Pankuweit S, Arbustini E, Basso C, Gimeno-Blanes J, Felix SB, Fu M, Heliö T, Heymans S, Jahns R, Klingel K, Linhart A, Maisch B, McKenna W, Mogensen J, Pinto YM, Ristic A, Schultheiss HP, Seggewiss H, Tavazzi L, Thiene G, Yilmaz A, Charron P, Elliott PM; European Society of Cardiology Working Group on Myocardial and Pericardial Diseases. Current state of knowledge on aetiology, diagnosis, management, and therapy of myocarditis: a position statement of the European Society of Cardiology Working Group on Myocardial and Pericardial Diseases. Eur Heart J. 2013;34:2636–2648, 2648a. doi: 10.1093/eurheartj/eht210. 2. Sagar S, Liu PP, Cooper LT Jr. Myocarditis. Lancet. 2012;379:738–747. doi: 10.1016/S0140-6736(11)60648-X. 3. Theleman KP, Kuiper JJ, Roberts WC. Acute myocarditis (predominately lymphocytic) causing sudden death without heart failure. Am J Cardiol. 2001;88:1078–1083. 4. Doolan A, Langlois N, Semsarian C. Causes of sudden cardiac death in young Australians. Med J Aust. 2004;180:110–112. 5. Aretz HT, Billingham ME, Edwards WD, Factor SM, Fallon JT, Fenoglio JJ Jr, Olsen EG, Schoen FJ. Myocarditis. A histopathologic definition and classification. Am J Cardiovasc Pathol. 1987;1:3–14. 6. Chow LH, Radio SJ, Sears TD, McManus BM. Insensitivity of right ventricular endomyocardial biopsy in the diagnosis of myocarditis. J Am Coll Cardiol. 1989;14:915–920. 7.Pinamonti B, Alberti E, Cigalotto A, Dreas L, Salvi A, Silvestri F, Camerini F. Echocardiographic findings in myocarditis. Am J Cardiol. 1988;62:285–291. 8. Mendes LA, Picard MH, Dec GW, Hartz VL, Palacios IF, Davidoff R. Ventricular remodeling in active myocarditis. Myocarditis Treatment Trial. Am Heart J. 1999;138(2 pt 1):303–308. 9. Hiramitsu S, Morimoto S, Kato S, Uemura A, Kubo N, Kimura K, Sugiura A, Itoh T, Hishida H. Transient ventricular wall thickening in acute 8 Steinl et al Ultrasound Molecular Imaging of Myocarditis Downloaded from http://circimaging.ahajournals.org/ by guest on May 12, 2017 myocarditis: a serial echocardiographic and histopathologic study. Jpn Circ J. 2001;65:863–866. 10. Kishimoto C, Hung GL, Ishibashi M, Khaw BA, Kolodny GM, Abelmann WH, Yasuda T. Natural evolution of cardiac function, cardiac pathology and antimyosin scan in a murine myocarditis model. J Am Coll Cardiol. 1991;17:821–827. 11. Narula J, Khaw BA, Dec GW, Palacios IF, Newell JB, Southern JF, Fallon JT, Strauss HW, Haber E, Yasuda T. Diagnostic accuracy of antimyosin scintigraphy in suspected myocarditis. J Nucl Cardiol. 1996;3:371–381. 12. Thavendiranathan P, Walls M, Giri S, Verhaert D, Rajagopalan S, Moore S, Simonetti OP, Raman SV. Improved detection of myocardial involvement in acute inflammatory cardiomyopathies using T2 mapping. Circ Cardiovasc Imaging. 2012;5:102–110. doi: 10.1161/CIRCIMAGING.111.967836. 13.Mahrholdt H, Goedecke C, Wagner A, Meinhardt G, Athanasiadis A, Vogelsberg H, Fritz P, Klingel K, Kandolf R, Sechtem U. Cardiovascular magnetic resonance assessment of human myocarditis: a comparison to histology and molecular pathology. Circulation. 2004;109:1250–1258. doi: 10.1161/01.CIR.0000118493.13323.81. 14.Francone M, Chimenti C, Galea N, Scopelliti F, Verardo R, Galea R, Carbone I, Catalano C, Fedele F, Frustaci A. CMR sensitivity varies with clinical presentation and extent of cell necrosis in biopsy-proven acute myocarditis. JACC Cardiovasc Imaging. 2014;7:254–263. doi: 10.1016/j. jcmg.2013.10.011. 15. Woodruff JF, Woodruff JJ. Involvement of T lymphocytes in the pathogenesis of coxsackie virus B3 heart disease. J Immunol. 1974;113:1726–1734. 16. Huber SA, Sartini D, Exley M. Vgamma4(+) T cells promote autoimmune CD8(+) cytolytic T-lymphocyte activation in coxsackievirus B3-induced myocarditis in mice: role for CD4(+) Th1 cells. J Virol. 2002;76:10785–10790. 17.Eriksson U, Ricci R, Hunziker L, Kurrer MO, Oudit GY, Watts TH, Sonderegger I, Bachmaier K, Kopf M, Penninger JM. Dendritic cellinduced autoimmune heart failure requires cooperation between adaptive and innate immunity. Nat Med. 2003;9:1484–1490. doi: 10.1038/nm960. 18. van Heeswijk RB, De Blois J, Kania G, Gonzales C, Blyszczuk P, Stuber M, Eriksson U, Schwitter J. Selective in vivo visualization of immune-cell infiltration in a mouse model of autoimmune myocarditis by fluorine-19 cardiac magnetic resonance. Circ Cardiovasc Imaging. 2013;6:277–284. doi: 10.1161/CIRCIMAGING.112.000125. 19.Villanueva FS, Lu E, Bowry S, Kilic S, Tom E, Wang J, Gretton J, Pacella JJ, Wagner WR. Myocardial ischemic memory imaging with molecular echocardiography. Circulation. 2007;115:345–352. doi: 10.1161/ CIRCULATIONAHA.106.633917. 20. Kaufmann BA, Lewis C, Xie A, Mirza-Mohd A, Lindner JR. Detection of recent myocardial ischaemia by molecular imaging of P-selectin with targeted contrast echocardiography. Eur Heart J. 2007;28:2011–2017. doi: 10.1093/eurheartj/ehm176. 21. Pummerer CL, Luze K, Grässl G, Bachmaier K, Offner F, Burrell SK, Lenz DM, Zamborelli TJ, Penninger JM, Neu N. Identification of cardiac myosin peptides capable of inducing autoimmune myocarditis in BALB/c mice. J Clin Invest. 1996;97:2057–2062. doi: 10.1172/JCI118642. 22. Lindner JR, Dayton PA, Coggins MP, Ley K, Song J, Ferrara K, Kaul S. Noninvasive imaging of inflammation by ultrasound detection of phagocytosed microbubbles. Circulation. 2000;102:531–538. 23. Lindner JR, Song J, Christiansen J, Klibanov AL, Xu F, Ley K. Ultrasound assessment of inflammation and renal tissue injury with microbubbles targeted to P-selectin. Circulation. 2001;104:2107–2112. 24. Gao S, Ho D, Vatner DE, Vatner SF. Echocardiography in mice. Curr Protoc Mouse Biol. 2011;1:71–83. doi: 10.1002/9780470942390.mo100130. 25. Bauer M, Cheng S, Jain M, Ngoy S, Theodoropoulos C, Trujillo A, Lin FC, Liao R. Echocardiographic speckle-tracking based strain imaging for rapid cardiovascular phenotyping in mice. Circ Res. 2011;108:908–916. doi: 10.1161/CIRCRESAHA.110.239574. 26.Raher MJ, Thibault H, Poh KK, Liu R, Halpern EF, Derumeaux G, Ichinose F, Zapol WM, Bloch KD, Picard MH, Scherrer-Crosbie M. In vivo characterization of murine myocardial perfusion with myocardial contrast echocardiography: validation and application in nitric oxide synthase 3 deficient mice. Circulation. 2007;116:1250–1257. doi: 10.1161/ CIRCULATIONAHA.107.707737. 27. Lindner JR, Song J, Xu F, Klibanov AL, Singbartl K, Ley K, Kaul S. Noninvasive ultrasound imaging of inflammation using microbubbles targeted to activated leukocytes. Circulation. 2000;102:2745–2750. 28.Felker GM, Boehmer JP, Hruban RH, Hutchins GM, Kasper EK, Baughman KL, Hare JM. Echocardiographic findings in fulminant and acute myocarditis. J Am Coll Cardiol. 2000;36:227–232. 29. McCarthy RE III, Boehmer JP, Hruban RH, Hutchins GM, Kasper EK, Hare JM, Baughman KL. Long-term outcome of fulminant myocarditis as compared with acute (nonfulminant) myocarditis. N Engl J Med. 2000;342:690–695. doi: 10.1056/NEJM200003093421003. 30. Cai YH, Alvarez A, Alcaide P, Duramad P, Lim YC, Jarolim P, Lowe JB, Luscinskas FW, Lichtman AH. Abrogation of functional selectin-ligand expression reduces migration of pathogenic CD8+ T cells into heart. J Immunol. 2006;176:6568–6575. 31. Weller GE, Lu E, Csikari MM, Klibanov AL, Fischer D, Wagner WR, Villanueva FS. Ultrasound imaging of acute cardiac transplant rejection with microbubbles targeted to intercellular adhesion molecule-1. Circulation. 2003;108:218–224. doi: 10.1161/01.CIR.0000080287.74762.60. 32. Kondo I, Ohmori K, Oshita A, Takeuchi H, Yoshida J, Shinomiya K, Fuke S, Suzuki T, Mizushige K, Kohno M. Leukocyte-targeted myocardial contrast echocardiography can assess the degree of acute allograft rejection in a rat cardiac transplantation model. Circulation. 2004;109:1056–1061. doi: 10.1161/01.CIR.0000115586.25803.D5. 33. Afanasyeva M, Georgakopoulos D, Rose NR. Autoimmune myocarditis: cellular mediators of cardiac dysfunction. Autoimmun Rev. 2004;3:476– 486. doi: 10.1016/j.autrev.2004.08.009. 34.Afanasyeva M, Georgakopoulos D, Belardi DF, Ramsundar AC, Barin JG, Kass DA, Rose NR. Quantitative analysis of myocardial inflammation by flow cytometry in murine autoimmune myocarditis: correlation with cardiac function. Am J Pathol. 2004;164:807–815. doi: 10.1016/ S0002-9440(10)63169-0. 35.Konishi M, Erdem SS, Weissleder R, Lichtman AH, McCarthy JR, Libby P. Imaging granzyme B activity assesses immune-mediated myocarditis. Circ Res. 2015;117:502–512. doi: 10.1161/ CIRCRESAHA.115.306364. 36. Eriksson U, Penninger JM. Autoimmune heart failure: new understandings of pathogenesis. Int J Biochem Cell Biol. 2005;37:27–32. doi: 10.1016/j. biocel.2004.06.014. CLINICAL PERSPECTIVE Myocarditis is an important cause for heart failure and sudden cardiac death, particularly in young individuals. Currently used cardiac tests for diagnosing myocarditis lack sensitivity or specificity. Ultrasound molecular imaging with site-targeted microbubble contrast agents detects inflammatory cell recruitment and endothelial activation. In this study, we examined whether contrast-enhanced ultrasound molecular imaging is feasible for the detection of specific components of the inflammatory process in myocarditis. In a murine model of autoimmune myocarditis, ultrasound molecular imaging with contrast agents targeted to leukocytes, to P-selectin on endothelial cells, and to CD4+ lymphocytes was performed together with advanced high-resolution ultrasound assessment of left ventricular function and histological evaluation. Signals from targeted contrast agents were elevated not only in histologically severe, but also in moderate myocarditis. Importantly, signal from targeted microbubbles in moderate myocarditis was not accompanied by a decrease in left ventricular function on highresolution 2-dimensional and strain imaging. Also, signal from CD4+ targeted microbubbles correlated to CD4+ lymphocytes in tissue. Our results suggest that ultrasound molecular imaging may be feasible for the detection and characterization of inflammation in patients with myocarditis even in the absence of functional impairment of the myocardium. Noninvasive Contrast-Enhanced Ultrasound Molecular Imaging Detects Myocardial Inflammatory Response in Autoimmune Myocarditis David C. Steinl, Lifen Xu, Elham Khanicheh, Elin Ellertsdottir, Amanda Ochoa-Espinosa, Martina Mitterhuber, Katharina Glatz, Gabriela M. Kuster and Beat A. Kaufmann Downloaded from http://circimaging.ahajournals.org/ by guest on May 12, 2017 Circ Cardiovasc Imaging. 2016;9: doi: 10.1161/CIRCIMAGING.116.004720 Circulation: Cardiovascular Imaging is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231 Copyright © 2016 American Heart Association, Inc. All rights reserved. Print ISSN: 1941-9651. Online ISSN: 1942-0080 The online version of this article, along with updated information and services, is located on the World Wide Web at: http://circimaging.ahajournals.org/content/9/8/e004720 Data Supplement (unedited) at: http://circimaging.ahajournals.org/content/suppl/2016/08/08/CIRCIMAGING.116.004720.DC1 Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in Circulation: Cardiovascular Imaging can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office. Once the online version of the published article for which permission is being requested is located, click Request Permissions in the middle column of the Web page under Services. Further information about this process is available in the Permissions and Rights Question and Answer document. Reprints: Information about reprints can be found online at: http://www.lww.com/reprints Subscriptions: Information about subscribing to Circulation: Cardiovascular Imaging is online at: http://circimaging.ahajournals.org//subscriptions/ SUPPLEMENTAL MATERIAL Supplementary table. Global and regional histology scores in animals with moderate (grade 1 and 2) and severe (grade 3 and 4) myocarditis. Global histology score Moderate myocarditis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Severe myocarditis 1 2 3 4 5 6 7 Regional histology scores Q1 Q2 Q3 Q4 2 2 2 1 1 2 1 1 2 2 2 1 1 1 1 0 0 0 1 0 0 1 2 0 0 1 1 1 0 2 2 0 1 2 1 0 2 0 2 0 0 0 2 2 2 1 1 2 0 1 2 0 0 0 0 0 1 0 1 0 1 0 0 0 1 2 0 0 0 0 3 3 4 4 3 4 4 1 2 4 4 3 4 4 2 3 4 4 3 4 4 3 2 4 4 3 4 4 3 1 4 4 3 4 4 Radial strain [%] 80 60 40 20 0 C Longitudinal strain [%] B Circumferential strain [%] A -40 -30 -20 -10 0 E 15 150 100 50 0 no myocarditis (n=20) myocarditis (n=21) Acoustic intensity (AU) Ejection fraction [%] D -25 -20 -15 -10 -5 0 †" 10 *" ‡" 5 0 no myocarditis myocarditis n=20 n=21 Supplementary Figure 1. Comparison of conventional echocardiography, strain measurements and molecular imaging between mice without and with with myocarditis. Radial (A), circumferential (B) and longitudinal strain (C), as well as ejection fraction (D) did not show differences between the groups. In contrast, molecular imaging (E) showed significantly elevated signals in animals with myocarditis for microbubbles targeted to CD4, P-selectin and to leukocytes (**p<0.01 vs MBIso. (***p<0.001 vs MBIso). Longitudinal strain 1.0 0.8 0.8 0.8 0.6 0.4 0.0 0.0 Sensitivity 1.0 0.6 0.4 0.2 0.2 0.2 0.4 0.6 0.8 0.0 0.0 1.0 0.6 0.4 0.2 0.2 0.4 0.6 0.8 0.0 0.0 1.0 0.4 0.6 1 - Specificity MBCD4 MBPSel MBLc 1.0 0.8 0.8 0.8 0.6 0.4 0.2 Sensitivity 1.0 0.6 0.4 0.4 0.6 1 - Specificity 0.8 1.0 0.0 0.0 0.8 1.0 0.6 0.4 Supplementary 0.2 0.2 0.2 0.2 1 - Specificity 1.0 0.0 0.0 1 - Specificity Sensitivity Sensitivity MBIso 1.0 Sensitivity Sensitivity Ejection fraction 0.2 0.4 0.6 1 - Specificity 0.8 1.0 0.0 0.0 0.2 0.4 0.6 0.8 1.0 1 - Specificity Figure 2. Receiver-operating –characteristic (ROC) curves for diagnosing myocarditis (n=20 mice without myocarditis, n=21 mice with myocarditis). Area under the curve (95% confidence interval) was 0.55 (0.35 – 0.7) for ejection fraction, 0.60 (0.41 – 0.78) for longitudinal strain, 0.57 (0.39 – 0.75) for MBCtr, 0.86 (0.74 – 0.98) for MBCD4, 0.90 (0.81 – 0.99) for MBPSel, and 0.82 (0.68 – 0.95) for MBLc.