Cardiopathogenic T Cells Govern Progression and Functional Remodeling in Inflammatory Cardiomyopathy and Chronic Myocarditis

Visual Abstract

Highlights

• •Acute T cell–mediated myocardial inflammation is either followed by smoldering chronic myocarditis or inflammatory cardiomyopathy.
• •Echocardiographic imaging mirrors the immunopathological processes underlying the progressive myocardial inflammatory disease.
• •Early impairment in cardiac strain predicts disease progression.

Summary

Echocardiographic imaging mirrors the immunopathological processes underlying the progression from acute myocarditis to inflammatory cardiomyopathy. T cell–mediated myocardial inflammation and edematous swelling in subepicardial regions at disease onset is followed by smoldering low-level myocardial inflammation or increasing cardiomyocyte loss and fibrotic remodeling. The resulting decline of cardiac function in the progressive disease is reflected and best predicted by impaired echocardiographic strain.

The accumulation of immune cells in the myocardium and inflammation-associated cardiomyocyte damage can cause various cardiac symptoms and disease states, including chest pain, arrhythmias, reduced myocardial function, and heart failure.1, 2, 3 According to current clinical definitions and guidelines, short-term myocardial inflammation lasting less than 1 month without coronary artery disease is classified as acute myocarditis (AM).^1,4 The degree of histopathological inflammatory changes can range from mild to severe, correlating with disease severity.³ Persistent accumulation of immune cells in the myocardium (>1 month) associated with altered myocardial function and/or morphology is generally referred to as inflammatory cardiomyopathy (iCMP).^4,5 However, the transition from AM to iCMP, including the sequence and degree of myocardial morphological and functional changes, remains poorly defined.

Episodes of AM are frequently triggered by viral infections, with most viruses acting via global immune activation, with or without direct infection of cardiomyocytes.6, 7, 8, 9, 10 Likewise, infections with other cardiotropic infectious agents such as Trypanosoma cruzi can lead to acute and long-lasting myocardial inflammatory disease.^11,12 The advent of immune checkpoint inhibitor treatment in cancer patients has highlighted the central role of T cells as primary drivers of myocarditis, demonstrating that direct cardiomyocyte injury, for instance by a cytopathic virus, is not a prerequisite for the development of myocardial inflammatory disease.^13,14 Treatment with antibodies targeting cell surface molecules regulating T-cell activity, such as programmed death and/or cytotoxic T-lymphocyte associated protein 4, can induce myocardial inflammation within weeks following the first administration.^13,15 Immune checkpoint inhibitor–induced myocarditis is a potentially severe complication with mortality rates reported as high as 50%^16,17 with self-reactive T cells being directed against the cardiac protein myosin heavy chain 6 (MYH6)^18,19 or other hitherto unknown antigens.²⁰ The presence of heart antigen-specific T cells in the circulation of healthy individuals^18,21 underscores that the balance between activation and restraint of autoreactive T cells is key to a better understanding of inflammatory myocardial disease.²² Therefore, elucidating the main mechanisms underlying T cell-mediated myocardial inflammation is essential for identifying the relevant clinical parameters and diagnostic modalities that enable early detection and prediction of the disease trajectory from AM to either clinically stable chronic myocarditis (CM) or progressive iCMP.

As echocardiography is the initial imaging modality used in clinical evaluation of myocarditis,²³ we first determined the range of echocardiographic functional and morphological alterations in a cohort of AM patients. To provide mechanistic insight into whether and to what extent changes in echocardiographic parameters reflect T cell–driven myocardial inflammatory disease, we used a T cell receptor transgenic mouse model, in which cardiopathogenic CD4⁺ T cells initiate AM and drive disease progression.24, 25, 26 We found that the early onset and the degree of cardiopathogenic CD4⁺ T cell infiltration into the myocardium critically determined the extent of cardiac remodeling and functional impairment. Antibody-mediated depletion of CD4⁺ T cells during the acute phase of myocarditis halted inflammation-induced changes in the myocardium, prevented cardiac fibrosis and halted progression to iCMP. Echocardiographic strain parameters were identified as the best predictors for progression to iCMP in mice. Collectively, the present study demonstrates that cardiopathogenic CD4⁺ T cells induce acute, yet reversible, inflammation-driven myocardial changes and that their persistence in cardiac tissue is a key factor driving functional cardiac remodeling.

Methods

The data supporting the findings of this study are available from the corresponding author upon request.

Study design and echocardiography

The ImmpathCarditis study is a prospective, exploratory study that recruits myocarditis patients at the University Hospital Zurich and the Cantonal Hospital St. Gallen. Patients with AM were included based on the following inclusion criteria: 1) ≥1 clinical symptom and ≥1 diagnostic criterion²; 2) confirmation of a typical late gadolinium enhancement pattern on cardiac magnetic resonance (CMR) imaging; 3) normal coronary angiography or coronary computed tomography scan (stenosis <50%); and 4) patients between the ages of 16 and 85 years. Patients were excluded from the study if they met the following exclusion criteria: 1) a diagnosis of any of the following autoimmune or chronic infectious diseases: systemic lupus erythematosus, Lyme disease, trypanosomiasis, HIV, hepatitis B virus, or hepatitis C virus; and 2) the presence of coronary artery disease with stenosis >50%. The study was conducted in accordance with the Declaration of Helsinki and approved by local ethics committees (BASEC 2021-01917). Written informed consent was obtained from all patients, and patients were considered ineligible if they were mentally or linguistically incapable of understanding the study procedure. Cardiac troponin T values used and reported in the analysis correspond to the peak troponin T serum levels measured in study patients during hospitalization (high-sensitivity cardiac troponin T Elecsys assay, Roche Diagnostics). For study patients recruited at the Cantonal Hospital St. Gallen, troponin T levels were measured in serum samples (high-sensitivity cardiac troponin T Elecsys assay, Roche Diagnostics) collected during the recruitment process (ie, 3-6 days after admission) and were used for analysis.

Echocardiography in AM patients was performed as part of the diagnostic work-up at the time of initial patient presentation. The transthoracic echocardiographic sequences were acquired using commercially available ultrasound systems (Epiq 7, Philips Medical Systems; E95, GE Healthcare). Acquisition and analysis of echocardiographic sequences was performed by experienced, certified personnel according to current guidelines.^27,28 Left ventricular ejection fraction (LVEF) was determined using the Simpson’s biplane method. Strain analysis was performed with TomTec Image Arena Cardiac Performance Analysis (version 4.6). Only echocardiographic sequences fulfilling the following quality criteria were considered for strain assessment: 1) available sequences in 2-, 3-, and 4-chamber and short-axis views; 2) sequences of at least 2 cardiac cycles with the endocardial margins clearly visible in all left ventricular (LV) segments; and 3) frame rate above 50 frames/s. Global strain measurements were based on the 16-segment LV model. Global longitudinal strain (GLS) was measured from apical view sequences and global circumferential strain (GCS) from short-axis views. Reference values for LV dimensions, function, and strain are based on previous recommendations of the American Society of Echocardiography and European Association of Cardiovascular Imaging and reported studies.^29,30

Experimental animals

All animal experiments were performed according to the federal and cantonal guidelines (Tierschutzgesetz) and were approved by the St. Gallen Cantonal Veterinary Office under the permission number SG/13/2023. The MYH6-specific T cell receptor transgenic mice (TCRM) on a BALB/c background were described previously.²⁴ TCRM mice were bred heterozygous, and transgene-negative littermates served as control subjects. Mice were treated with 500 μg of anti-CD4 antibody (clone: YTS191, ECACC, 87072282) or IgG2b isotype control antibody (clone: MPC-11, ECACC, ATCC CCL-167) by intraperitoneal injection twice a week as demonstrated in the respective experiments.

Echocardiographic analysis in mice

Assessment of cardiac function and morphology was performed with transthoracic echocardiography using the Vevo 3100 ultrasound machine (VisualSonics Inc) and analyzed using Vevo LAB version 5.7.1 (VisualSonics Inc). Mice were sedated initially with 4% isoflurane (Attane, Provet) and compressed breathing air, after induction mice were transferred on a heated plate (maintenance of body temperature at 37 °C) and the fur was removed over the thorax using hair removal cream (Veet PURE). Throughout the procedure, anesthesia was maintained with 1% to 1.5% isoflurane. Vital parameters were monitored constantly, heart rate was kept between 400 and 500 beats/min. Body temperature was measured with a rectal temperature sensor and was maintained between 36.5 °C and 37.5 °C. Evaluation of LV function was measured in B-mode images in the parasternal long-axis and parasternal short-axis view. LV dimensions were assessed in M-mode images in short-axis view. All functional and dimensional parameters were analyzed in 3 different sequences, and the calculated mean was used for further analysis. Global longitudinal strain was measured in the long-axis view. Global circumferential and radial strain were assessed in the short-axis view. Strain parameters were analyzed with the Vevo strain analysis software on 3 cardiac cycles.

Flow cytometry

Flow cytometry of cardiac immune cell content was performed as previously described.²⁶ In brief, sacrificed mice were perfused with 20 mL of phosphate-buffered saline (PBS) (Fisher BioReagents) before organ harvest. Hearts were minced into small pieces and placed into a 6-well dish filled with RPMI 1640 medium (PAN-Biotech) containing 2% fetal bovine serum (FBS) (Capricon Scientific), 1 mg/mL collagenase P (Sigma-Aldrich), and 25 μg/mL DNase I (Applichem), and incubated at 37 °C under continuous agitation (100 rpm) for 1 hour. Residual tissue fragments were then mechanically disrupted. For immune cell analysis, mononuclear cells were enriched by centrifugation (25 minutes at 800 g, 4 °C) on a 30% to 70% Percoll gradient (Cytiva). LIVE/DEAD cell discrimination of single-cell suspensions was performed by using a fixable BV510 Zombie Aqua viability dye (BioLegend) in PBS for 20 minutes at 4 °C. After washing, cells were incubated for 20 minutes at 4 °C in PBS-containing 2% FBS and 10 mmol/L ethylenediaminetetraacetic acid with the following fluorochrome-labeled antimouse antibodies: CD45 (BU496), CD4 (BUV805), Vα2 TCR (APC), Vβ8.1/8.2 TCR (FITC), CD3 (A700), CD11b (BUV395), and Ter119 (BV510) (Supplemental Table S1). Cells were acquired with the BD FACSSymphony A3 (BD Biosciences) and analyzed using FlowJo version 10.6.2 software (FlowJo LLC) following established guidelines.³¹

Quantification of peripheral lymphocytes was performed by flow cytometry. The 2 to 3 drops of blood were suspended in PBS containing 2% FBS and 10 mmol/L ethylenediaminetetraacetic acid. Cells were incubated for 20 minutes at 4 °C with the following fluorochrome-labeled antimouse antibodies: CD45 (APC C7), CD3 (PeCy7), CD4 (APC), Vβ8.1/8.2 TCR (FITC), and CD8α (PE) (Supplemental Table S1). Erythrocyte lysis was performed by adding 1 mL of FACS lysing solution (BD Biosciences, 1:10). Samples were measured using the BD LSRFortessa (BD Biosciences) and analyzed with FlowJo version 10.6.2 software (FlowJo LLC).

Histopathological analysis of myocardial inflammation and fibrosis

Histopathological analysis was performed as previously described.²⁴ Hearts were cut axially (base, mid, and apex part) and fixed with 4% formaldehyde (Formafix, Biosystems) for at least 12 hours and embedded in paraffin. Histopathological changes were evaluated using a semiquantitative scoring system for myocarditis severity based on hematoxylin and eosin staining: 0, no inflammation; 1, <100 inflammatory cells involved, small inflammatory lesions; 2, >100 inflammatory cells involved, larger inflammatory lesions; 3, >10% of the heart section involved in inflammation; 4, >30% of the heart section involved in inflammation; and 5, >30% of the heart section involved in inflammation with extensive fibrosis and dilation of ventricle. Cardiac fibrosis was determined with a picrosirius red staining kit according to the manufacturer’s instructions (Cardiac Muscle, Abcam). All histopathological analyses were conducted on consecutive tissue sections and in a blinded manner. Images from heart sections were scanned with a Pannoramic 250 Flash III scanner547 (3DHISTECH) at the Institute of Pathology, Cantonal Hospital St. Gallen. Cardiac fibrosis was assessed by quantifying the picrosiruis red-positive area in the mid and the apex part using the Qupath software (version 04.1).

Statistical analysis

Continuous data are presented using actual data points with the mean ± SD, median (IQR), or range as specified. Categorical data are presented as counts and percentages. The Shapiro-Wilk test was used to determine if data were normally distributed. The deviation of data points from the respective reference ranges of human clinical data was assessed using the 1 sample Wilcoxon signed-rank test, with the median of each reference range serving as comparator. Comparisons between 2 groups used Student’s t-test or Mann-Whitney U test for normal and skewed data, respectively. Comparisons among >2 groups used 1-way analysis of variance with Dunnett's or Tukey's post hoc test for multiple pairwise comparisons or Kruskal-Wallis test for non-normally distributed data. Pearson's or Spearman's correlation coefficients were used to evaluate associations between continuous variables dependent on data distribution. Survival analysis was performed using Kaplan-Meier curves and the log-rank test. A receiver-operating characteristic (ROC) analysis was performed using logistic regression to determine if LVEF or GLS could distinguish CM from iCMP with results resented as the area under the ROC curve (AUC) with 95% CIs.

Statistical analyses were performed using Graphpad Prism version 10.0 (Graphpad Software, Inc), and statistical significance was defined as a P value <0.05

Results

Echocardiographic alterations in patients with AM

To determine the range of echocardiographic alterations in AM patients at first clinical presentation and to establish clinically relatable parameter sets for subsequent mechanistic studies in a preclinical model, we enrolled 58 AM patients in the prospective ImmpathCarditis study. The majority of the AM patients were young men (Table 1), which is comparable to the median age and sex distribution observed in other cohorts of the disease.³² Acute infection symptoms were reported by 84% within 4 weeks before admission. However, in 57% of these cases, no infectious agent could be detected in serological analysis using PCR. Given that the majority of myocarditis patients presented more than 7.5 days after symptom onset of infection, it is plausible that the infectious trigger had already been cleared in some individuals by the time of testing. These findings support the growing body of evidence suggesting that preceding infections may act primarily as a trigger of autoimmune responses rather than serving as the direct etiological cause of myocarditis.^21,33 AM was diagnosed based on positive Lake Louise Criteria³⁴ in CMR (Table 1). All patients demonstrated cardiac troponin T levels above the 99th percentile of cardiac healthy individuals in central Europe³⁵ (Table 1, Figure 1A). Alterations in the indexed left ventricular end-systolic volume (LVESV) and indexed left ventricular end-diastolic volume (LVEDV) were shown by 39.2% and 13.4% of the AM patients, respectively (Figures 1B and 1C). The resulting LVEF was reduced in 39.7% of patients with 52% LVEF considered as lower reference value³⁶ (Figure 1D). Impaired myocardial deformation was detected in 68.4% of patients with GLS lower than −18% and 56.9% of patients with GCS lower than −23% (Figures 1E and 1F). Moreover, 5% to 10% of the AM patients showed alterations in morphological parameters in echocardiography (Supplemental Figure S1). LVEF values were significantly inversely correlated with other functional parameters including indexed LVESV, GLS, and GCS (Figures 1G and 1H). Serum concentration of troponin T showed a significant inverse correlation with LVEF and a positive correlation with indexed LVESV and LVEDV (Figures 1G and 1H). In sum, these data underscore that acute cardiac inflammation can lead to profound functional and morphological changes in the myocardium, with a substantial proportion of patients showing acute heart failure-like changes in key echocardiographic parameters.

Table 1.

Characteristics of Patients With Acute Myocarditis^a

Age, y	27 (17-66)
Male/female	50/8 (86)
History
Previous myocarditis	10 (17)
Previous acute infectionb	49 (84)
Onset acute infection prior admission, d	7.5 (3.0-33.0)
Symptoms of acute infection
Gastrointestinal	7/49 (14)
Respiratory	31/49 (63)
Other infection	11/49 (22)
Cause of acute infection
Bacterial	9/49 (18)
Viral	12/49 (24)
Unknown	28/49 (57)
Diagnostic parameters
CMR positivec	58 (100)
LV wall motion irregularitiesd	30/58 (52)
CRP, mg/L	53 (1-535)
Troponin T, ng/L	1,369 (57-9,425)
NT-proBNP, ng/L	832 (30-32,392)

Values are median (range) or ratio (%).

CRP = C-reactive protein; NT-proBNP = N-terminal pro–B-type natriuretic peptide.

^aPatients (n = 58) were recruited at the University Hospital Zurich or the Cantonal Hospital St. Gallen, Switzerland.

^bOccurrence of acute infection within 4 weeks before clinical presentation were recorded.

^cCardiac magnetic resonance (CMR)–confirmed acute myocarditis according to the Lake Louise criteria.

^dLeft ventricular (LV) wall motion irregularities detected on echocardiography.

Figure 1.

Morphological and Functional Echocardiographic Parameters in Acute Myocarditis Patients

(A) Peak troponin T serum levels in patients with acute myocarditis during hospitalization. (B to F) Morphological and functional parameters of the left ventricle at the initial echocardiographic assessment on hospital admission: (B) left ventricular end-systolic volume indexed (LVESVi), (C) left ventricular end-diastolic volume indexed (LVEDVi), (D) left ventricular ejection fraction (LVEF), (E) global longitudinal strain (GLS), (F) global circumferential strain (GCS). (G and H) Correlation analysis between troponin T serum concentration and echocardiographic parameters. (A to F) Dots represent values of individual patients; box and whiskers show minimum to maximum, median and IQR. Dotted lines represent reference values of clinical parameters according to guidelines. Bar plots represent percentage of patients with parameters within (gray) and outside (red) reference values. Reference range B: 9-31 mL/m²; C: 34-74 mL/m²; D: 52%-72%; E: −18% to −27.7%; F: −23% to −40.6%; A to F: 1-sample Wilcoxon signed-rank test was used. G and H: Statistical analysis was performed using nonparametric Spearman correlation analysis. ∗P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001. ns = nonsignificant.

2 distinct disease trajectories in T Cell-Mediated myocarditis

To elaborate the temporal changes in functional and morphological cardiac parameters in a defined preclinical setting of AM with progression to a more severe disease, we resorted to a T-cell receptor transgenic mouse model in which more than 90% of the CD4⁺ T cells recognize the immunodominant MYH6_614–629 epitope (TCRM).^24,25 In a longitudinal cohort of 24 TCRM mice followed for 12 weeks, 33% of the animals developed overt clinical disease (ie, apathy and dyspnea) beginning at 6 to 8 weeks of age (Figure 2A). Longitudinal high-resolution echocardiography revealed substantial morphological changes in 4-week-old TCRM mice in comparison to nontransgenic littermates (Supplemental Figures S2B to S2G), including a significant reduction in the systolic left ventricular internal diameter (LVID ^systole) (Figure 2B). Likewise, 4-week-old TCRM mice showed significant alterations in functional echocardiographic parameters, with an increased LVEF (Figure 2C), reduced LVESV and LVEDV (Supplemental Figures S2H and S2I), and an enhanced circumferential myocardial shortening, as indicated by a decreased GCS (Figure 2D). In 6-week-old TCRM mice, longitudinal myocardial shortening was significantly reduced, as evidenced by an impaired GLS compared with control animals (Figure 2E). In contrast, global radial strain appeared to be only mildly affected throughout the disease course (Supplemental Figure S2J). Decreased LVEF and impaired GCS and GLS in some 10- to 12-week-old TCRM mice indicated that ongoing inflammation affected heart function in these animals (Figures 2C to 2F). Gross pathological examination indicated that mice with severe immune cell infiltration and increased fibrosis developed ventricular dilation (Figure 2G). The progression to advanced disease in the TCRM mice led to an attrition of the cohort to 66% (16 of 24) after 10 weeks (Figure 2A and 2B). Histopathological analysis at 12 weeks of age or upon premature termination caused by overt symptoms revealed either mild (myocarditis score ≤3) or severe disease (myocarditis score ≥4) (Figure 2H, Supplemental Figure S2A). Mice with more severe immune cell infiltration showed significantly increased collagen deposition (≥20%), as determined by picrosirius red staining (Figures 2H and 2I). Indeed, histopathologically severe myocardial disease (defined as myocarditis score ≥4 and collagen deposition ≥20%) correlated with significantly lower LVEF and impaired GCS and GLS when compared with mild disease (Supplemental Figures S2K to S2M). Comparing these key functional parameters at the respective endpoints in TCRM mice to healthy control subjects confirmed the distinction of mild and severe disease (Figure 2J). The significant functional and morphological alterations in TCRM hearts with severe fibrosis and high inflammation indicated that AM had developed into iCMP, whereas TCRM hearts with mild fibrotic remodeling and no/low functional impairment had maintained the state of CM (Figure 2J). Moreover, these findings suggest that the disease trajectory with either progression from AM to iCMP or the sustenance of clinically stable CM is determined shortly after disease onset.

Figure 2.

Temporal Sequence of Morphological and Functional Echocardiographic Parameters During Myocarditis

(A) Survival analysis of T-cell receptor transgenic mice (TCRM) and control mice (Ctrl). (B to E) Dynamics of echocardiographic parameters in TCRM vs Ctrl mice. (B) Left ventricular internal diameter systole (LVID systole), (C) LVEF, (D), GCS, (E), GLS. (F) Representative echocardiographic sequences of the left ventricle in short axis (SAX) and motion-mode (M-mode) view from TCRM and Ctrl mice at 4 and 8 weeks of age. (G) Gross pathology at endpoint. (H) Representative images of hematoxylin and eosin (H&E) and picrosirius red (PSR)-stained hearts of TCRM mice with distinct disease courses and Ctrl at endpoint (12 weeks of age or premature termination). (I) Quantification of PSR-positive area. (J) Fold change in functional parameters of TCRM mice with inflammatory cardiomyopathy (iCMP) (ie, severe) or with chronic myocarditis (CM) (ie, mild) compared with Ctrl. (B) Fraction of TCRM mice still alive at the respective time point. (A to J) Pooled data from 3 independent experiments [n (Ctrl) 19; n (TCRMmild) = 9; n(TCRMsevere) = 15]. (C) n(TCRMsevere) = 12 (3 mice excluded from histopathologic analysis caused by rapid deterioration of disease). Dots represent individual mice; box and whiskers show minimum to maximum, median and IQR. (F to I) Violin plots show median and IQR. Statistical analyses used: (A) log-rank test, (B to E) unpaired 2-tailed Student’s t-test or Mann-Whitney U test. (I) 1-way analysis of variance with Dunnett`s multiple comparisons test. ∗P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001. Abbreviations as in Figure 1.

Early changes in myocardial strain predict myocarditis outcome

To determine to what extent the key functional echocardiographic parameters predicted disease outcome in TCRM mice, we performed a subgroup analysis based on retrospective stratification according to the disease phenotype, ie, CM vs iCMP with significantly increased collagen deposition (Figure 2C) and decreased survival in the iCMP group (Supplemental Figure S3A). Time-resolved LVEF analysis revealed a continuous decline in cardiac function in mice with iCMP, whereas LVEF in the CM group remained stable (Figure 3A). The 4-week-old TCRM mice of both groups showed significantly elevated LVEF when compared with control subjects (Figure 3B), whereas at the 6-week time point LVEF was significantly lower in iCMP vs CM mice (Figure 3C). Interestingly, the early changes in the LVEF were only a weak predictor for progression to iCMP as shown in the ROC curve with an area under the curve (AUC) of 0.603 (Figure 3D). Likewise, the minor differences in morphological parameters such as systolic LVID ^systole at the 6-week time point were not predictive for the progression to iCMP (Supplemental Figures S3B to S3E). The 4-week-old TCRM mice in both groups showed reduced GCS in comparison with control subjects (Figures 3E and 3F) and an impaired GCS associated with the iCMP phenotype at the age of 6 weeks (Figure 3G). The reduced GCS at 4 weeks and an AUC of 0.726 in the iCMP group (Figure 3H) support the notion that the degree of early myocardial inflammation determines the long-term consequences of the disease. GLS showed a continuous decline in the iCMP group (Figure 3I) similar to the other key functional parameters LVEF (Figure 3A) and GCS (Figure 3E). Although GLS was not affected by early myocardial inflammation at 4 weeks of age (Figures 3I and 3J), significant changes appeared at the 6-week time point with impairment in the iCMP group (Figure 3K). The ROC curve analysis revealed that GLS showed a highly significant diagnostic accuracy for iCMP progression with an AUC of 0.915 (Figure 3L). Taken together, these data show that early changes in echocardiographic strain serve as accurate indicators for the progression of AM to iCMP.

Figure 3.

Predictive Value of Functional Echocardiographic Parameters on Myocarditis Outcome

Alterations in myocardial function during myocarditis, stratified by disease outcome comparing mice with CM or iCMP to Ctrl. (A to D) Changes in LVEF in different disease outcomes: (A) time course analysis, (B-C) critical time point analysis at B, 4 weeks and C, 6 weeks of age. (D) ROC curve for LVEF distinguishing iCMP from Ctrl at 6 weeks of age. (E to H) Changes in GCS in different disease outcomes: (E) time course analysis, (F and G) critical time point analysis at F, 4 weeks and G, 6 weeks of age. (H) ROC curve for GCS distinguishing iCMP from Ctrl at 4 weeks of age. (I to L) Changes in GLS in different disease outcomes: I, time course analysis, J-K, critical time point analysis at J, 4 weeks and K, 6 weeks of age. (L) ROC curve for GLS distinguishing iCMP from Ctrl at 6 weeks of age. (A, E, I) Dots represent the mean of each subgroup per time point, whiskers show ± SD. (B, G) Kruskal-Wallis test with Dunnett’s multiple comparison; (C, F, J, K) ordinary 1-way analysis of variance with Tukey’s multiple comparison. AUC = area under the curve; ROC = receiver-operating characteristic. Boxplots as in Figure 2. ∗P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001. Abbreviations as in Figures 1 and 2.

Immune cell infiltration is associated with rapid-onset myocardial remodeling

To assess how early functional changes in TCRM hearts track with immune cell infiltration and fibrosis, we used flow cytometric analysis of heart-infiltrating immune cells or quantitative histopathological assessment, each in combination with echocardiography (Figure 4A, Supplemental Figure S4A). Flow cytometry-based immune cell quantification at the age of 4, 6, and 8 weeks revealed significantly increased accumulation of CD45⁺ immune cells (Figure 4B), MYH6-specific CD4⁺ T cells expressing the transgenic T cell receptor (TCR) Vα2 and Vβ8 chains (Figure 4C, Supplemental Figure S4A), and myeloid cells expressing the marker CD11b (Figure 4D, Supplemental Figure S4A) compared with transgene-negative littermate controls. Histopathological analysis of the second set of mice showed that immune cells initially accumulated mainly in subepicardial regions (Figure 4E, arrows) before pinching into the cardiac parenchyma in 6-week-old TCRM mice (Figure 4E, arrowheads), leading to increased cardiomyocyte damage at the 8-week time point (Figure 4E, asterisks). The rapid onset and the progressive immune cell infiltration, as indicated by the rising myocarditis score (Figure 4F), was accompanied by enhanced collagen deposition (Figure 4G). Notably, collagen deposition was particularly strong in areas of cardiomyocyte damage and loss (Figure 4E, double arrows). Impairment in echocardiographic parameters was consistent with the longitudinal analysis (Figure 2), showing significant changes in LVID ^systole (Figure 4H), LVEF (Figure 4I), and GCS (Figure 4J) at the 4-week time point, along with significant impairment of GLS in hearts of 6-week-old TCRM mice (Figure 4K). Multiple correlation analysis revealed that the early immune cell accumulation of 4-week-old TCRM hearts precipitated significant morphological changes and mostly affected the GCS (Figures 4L and 4M). Progressive fibrosis in 6- to 8-week-old TCRM hearts showed a significant positive correlation with LVID ^systole, GLS, and GCS and a significant negative correlation with LVEF (Figures 4N and 4O). In sum, these data indicate that immune cell infiltration leads to rapid remodeling of the myocardium that is reflected by alterations in morphological and functional echocardiographic parameters.

Figure 4.

Correlation Between Immune Cell Infiltration, Echocardiographic Alterations, and Cardiac Remodeling

(A) Experimental design. (B to D) Heart-infiltrating immune cells in TCRM and control mice at 4, 6, and 8 weeks of age measured by flow cytometry: B, CD45⁺ leucocytes, C, CD4⁺ T cells, D, CD11b⁺ myeloid cells. (E) Representative heart sections stained with H&E and PSR. (F) Myocarditis score and (G) quantification of PSR-positive stained area in mice at 4, 6, and 8 weeks of age. (H to K) Echocardiographic assessment of left ventricular dimension and function in 4-, 6-, and 8-week-old TCRM and control mice. (H) LVID systole, (I) LVEF, (J) GCS, (K) GLS. (L and M) Correlation matrix showing the relationship between immune cells and echocardiographic parameters in 4-week-old TCRM mice, with corresponding P values. (N and O) Correlation matrix showing the relationship between collagen deposition and echocardiographic parameters in 6- and 8-week-old TCRM mice, with corresponding P values. (B to D, L and M) Flow cytometry: Pooled data of 8 independent experiments including 4-week [n(TCRM) = 8, n(Ctrl) = 5], 6-week [n(TCRM) = 12, n(Ctrl) = 10], and 8-week-old mice [n(TCRM) = 12, n(Ctrl) = 5]. (E to G, N and O) Histology: Pooled data of 6 independent experiments: 4-week [n(TCRM) = 8, n(Ctrl) = 11], 6-week [n(TCRM) = 7, n(Ctrl) = 7] and 8-week-old mice [n(TCRM) = 13, n(Ctrl) = 9]. (H to K) Echocardiographic analysis of all mice. (B to G) Statistical analysis was performed using Kruskal-Wallis test with Dunnett`s multiple comparisons test. (H to K) Unpaired 2-tailed Student’s t-test or Mann-Whitney U test and (L and O) parametric Pearson correlation or nonparametric Spearman correlation analysis. Boxplots as in Figure 2. ∗P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001. Abbreviations as in Figures 1 and 2.

Myocardial functional remodeling is mediated by cardiopathogenic CD4⁺ T cells

Different immune cells have been shown to drive and/or to participate in myocardial inflammatory disease including macrophages, neutrophils, and T cells.^14,37, 38, 39 The first wave of cardiopathogenic T cells induce AM in TCRM mice before the age of 4 weeks^24,25 with detectable changes in myocardial morphology and function. Importantly, at this early stage, widespread cardiomyocyte loss and fibrotic remodeling is not histologically evident (Figures 4E to 4G). To assess the effect of T cell-driven inflammation on morphology and function at later stages of disease progression, we treated 4-week-old TCRM mice with an anti-CD4 antibody at the critical time window before the onset of irreversible fibrotic remodeling. Antibody treatment resulted in a >99% reduction in CD4⁺ T cell counts in the circulation (Figures 5A to 5C) compared with TCRM mice treated with an isotype control antibody. Before treatment, baseline functional parameters in TCRM mice were recorded by echocardiography, confirming a reduction in LVID ^systole (Supplemental Figure S5B), increase in LVEF (Supplemental Figure S5C), decrease in GCS (Supplemental Figure S5D), and stable GLS (Supplemental Figure S5E) compared with 4-week-old littermate control subjects. Histopathological analysis after 4 weeks of antibody treatment validated the dichotomy of mild vs severe disease phenotype in isotype control antibody-treated TCRM mice (Figures 5D to 5F). Depletion of CD4⁺ T cells reduced collagen deposition (Figure 5E) to levels comparable to littermate controls (Figures 2I and 4G), while the myocarditis score, ie, immune cell infiltration and cardiomyocyte damage as assessed by hematoxylin and eosin staining, remained elevated (Figure 5F). Echocardiographic assessment at 8 weeks of age corroborated the clear distinction of severe vs mild cardiac dysfunction in isotype control antibody-treated TCRM mice (Figures 5G to 5J). Depletion of CD4⁺ T cells reversed initial inflammation-induced changes in morphology and significantly improved cardiac function when compared with both mildly and severely affected isotype control antibody-treated TCRM mice (Figures 5G to 5J). Collectively, these data underscore that initial inflammation-induced morphological and functional alterations during AM are reversible, and that the persistence of cardiopathogenic CD4⁺ T cells within the myocardium is essential for myocardial functional remodeling and consequently the progression of acute myocardial inflammation to iCMP and the sustenance of CM.

Figure 5.

The Role of CD4⁺ T Cells in Cardiac Functional Remodeling

(A) Experimental design. (B and C) Flow cytometric analysis of CD4⁺ T cells in peripheral blood of TCRM mice treated with IgG2b isotype control or anti-CD4 antibody. (D) Representative H&E and PSR-stained heart sections. (E) Myocarditis score and (F) quantification of PSR-positive area, in 8-week-old TCRM mice subjected to the indicated treatment. (G to J) Echocardiographic assessment of left ventricular morphology and function in 8-week-old TCRM mice treated with isotype control or anti-CD4 antibody. (G) LVID systole, (H) LVEF, (I) GCS, (J) GLS. (C, E to J) Pooled data from 3 independent experiments, including n = 12 anti-CD4-treated TCRM mice and n = 13 isotype-treated TCRM mice of which n = 5 progressed to iCMP. Statistical analysis was performed using C, unpaired 2-tailed Student’s t-test or Mann-Whitney U test and (E to J) 1-way analysis of variance with Tukey’s post hoc test or Kruskal-Wallis test with Dunnett’s multiple comparisons test. Boxplots as in Figure 2. ∗P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001. Abbreviations as in Figures 1 and 2.

Discussion

The homeostatic heart is composed of various cell types, including fibroblasts that cater to cardiomyocytes and tissue-resident immune cells such as macrophages that provide additional functional support for cardiomyocytes.^26,40,41 Inflammatory perturbation, such as the infiltration of autoimmune T cells into the myocardium, leads to profound cardiomyocyte damage,^21,26,42 as reflected by the substantial troponin T release and the significant negative correlation between LVEF and troponin T in the cohort of AM patients analyzed here. The time-resolved echocardiographic analysis of CD4⁺ T cell-driven autoimmune myocarditis performed in this study revealed that early immune cell accumulation in subepicardial regions of TCRM hearts (ie, at the age of 4 weeks) was associated with increased LVEF and decreased LVID and GCS. These findings are consistent with the view that myocardial dysfunction, including wall motion abnormalities, in patients with AM is associated with edematous swelling of the myocardium.⁴³ The sustained influx of immune cells into the hearts of 6- to 8-week-old TCRM mice resulted in increased cardiomyocyte loss and enhanced collagen deposition, which precipitated profound changes in GLS, a finding that mirrors the positive correlation between scar formation and GLS in patients with AM.⁴⁴ Thus, the following 2-stage scenario for early cellular processes and functional consequences in the acutely inflamed myocardium emerges: first, initial T cell-mediated myocardial inflammation leads to edematous swelling mainly in subepicardial regions with limited cardiomyocyte damage and fibrosis, as reflected by increased LVEF and decreased GCS. In the second stage, prolonged and more widespread immune cell accumulation in the myocardium causes increased cardiomyocyte loss and extensive fibrotic remodeling, resulting in a decline of cardiac function, with GLS being the first functional parameter to be impaired.

Regional or scattered accumulation of immune cells in the myocardium is considered one of the hallmarks of iCMP^4,5 and serves as a valuable clinical phenotypical trait for the subclassification of dilated cardiomyopathy^45,46 or arrhythmogenic right ventricular cardiomyopathy.⁴⁷ However, the distinction of CM as a persisting myocardial inflammation without major histologically detectable cardiomyocyte damage and hence limited fibrosis remains challenging.^4,48 Notably, myocardial inflammation has been detected as incidental postmortem finding in the absence of clinically overt cardiac disease^49,50 suggesting that subclinical myocardial inflammation can persist without being diagnosed. The findings in TCRM mice with persisting CM support this notion, namely that a distinct threshold of myocardial immune cell accumulation has to be overcome to cause overt clinical disease. In CD4⁺ T cell-driven myocardial inflammation of TCRM mice, the disease trajectory from AM to either CM or the severe clinical disease, ie, iCMP, was determined both by the degree of immune cell infiltration and the extent of fibrotic remodeling. Sustenance of the subclinical CM in TCRM mice was accompanied by a persistent moderate increase in LVEF and a decrease in GCS, which can be interpreted as ongoing edematous myocardial swelling mainly in the subepicardial region. Although the prognostic value of increased CMR strain parameters at initial diagnosis for the progression of AM patients to a more severe disease has been demonstrated,⁵¹ systematic cardiac imaging studies monitoring the functional and morphological changes in CM and/or recurrent AM are still missing.

The extended cardiac imaging analysis in the preclinical TCRM model provided here demonstrates the delicate balance between vulnerability and resilience of the myocardium to immune cell-mediated damage. The clear dichotomy of AM either progressing to iCMP with extensive myocardial damage and fibrosis vs the more benign disease course in the AM-to-CM transition underscores that different treatment options for such distinct conditions need to be considered. Under conditions of severe disease with strong T cell-driven myocardial damage, immunosuppressive intervention with the aim to reduce mainly the activity of T cells^52,53 is certainly warranted. Whether and to what extent additional disease-modulating approaches such as the attenuation of proinflammatory pathways⁵⁴ or preservation of cardiac fibroblast homeostasis²⁶ will be suitable for the treatment of acute-severe vs chronic-mild disease conditions requires further clinical phenotyping and mechanistic dissection.

In sum, the present study underscores that the extent of the initial cardiopathogenic CD4⁺ T cell infiltration into the myocardium determines whether the disease progresses to iCMP with heart failure or remains at the state of a chronically smoldering myocardial inflammation. The finding that the degree of myocardial remodeling and functional impairment is directly linked to regional T cell accumulation in the heart provides a conceptual framework for improving diagnostic precision and distinguishing clinical phenotypes in myocarditis.

Study limitations

Our study highlights the diagnostic potential of echocardiographic strain parameters in detecting early myocardial changes and predicting the disease outcome in a preclinical model of T cell-driven myocardial inflammation. The clinical confirmation and implementation of these findings requires well-designed, prospective clinical trials in patients with suspected or confirmed myocarditis. Second, the immunological factors contributing to the high disease variability both in the genetically homogenous TCRM mice and in myocarditis patients require further investigation. However, it is not possible to infer from the TCRM model that CD4⁺ T cell depletion is a directly translatable strategy for humans. Instead, immunomodulatory approaches should concentrate on the selective targeting of cardiopathogenic T cells. The cellular processes and molecular mechanisms underlying reversible myocardial injury, as demonstrated by CD4⁺ T cell depletion, remain unclear and require future studies that combine immune cell phenotyping and high-resolution spatial analysis of the inflamed myocardium.

Conclusions

Our study highlights that T cell-mediated myocardial inflammation and edematous swelling in subepicardial regions at disease onset is either followed by smoldering low-level myocardial inflammation or increasing cardiomyocyte loss and fibrotic remodeling. Echocardiographic imaging mirrors the immunopathological processes underlying the progression from AM to inflammatory cardiomyopathy. The decline of cardiac function in the progressive disease is reflected and best predicted by impaired echocardiographic strain.

Perspectives.

COMPETENCY IN CLINICAL KNOWLEDGE: Echocardiographic strain analysis may serve as a sensitive, noninvasive tool for early risk stratification in patients with AM and subsequent clinical follow-up monitoring.

TRANSLATIONAL OUTLOOK: This study demonstrates that cardiopathogenic CD4⁺ T cells induce acute, yet reversible, inflammation-driven myocardial changes and that their persistence in cardiac tissue is a key factor driving functional cardiac remodeling. Thus, targeting pathogenic T cell responses could offer a therapeutic strategy to prevent progression from acute myocardial inflammation to inflammatory cardiomyopathy.

Funding Support and Author Disclosures

This study was financially supported by the Swiss National Science Foundation (grant 10000830 to Dr Ludewig), the European Research Council (Advanced Grant contract 101019872, CardiacStroma, to Dr Ludewig), the Promedica Foundation (1704/M to Dr Ludewig) and the Swiss Heart Foundation (FF23053 to Dr Ludewig). Drs Gil-Cruz, Perez-Shibayama, and Ludewig are founders and shareholders of Stromal Therapeutics; and are listed as inventors on patent WO 2022/084400 A1. Dr Ruschitzka has not received personal payments by pharmaceutical companies or device manufacturers in the last 3 years (remuneration for the time spent in activities, such as participation as steering committee member of clinical trials and member of the Pfizer Research Award selection committee in Switzerland, were made directly to the University of Zurich); the Department of Cardiology (University Hospital of Zurich/University of Zurich) reports research, educational, and/or travel grants from Abbott, Abiomed, Alnylam, Amarin, Amgen, AstraZeneca, At the Limits Ltd, Bayer, Biotronik, Bristol Myers Squibb, Boehringer Ingelheim, Boston Scientific, Bracco, CM Microport, Concept Medical, CTI, Daiichi Sankyo, Davos Congress, Edwards Lifesciences, FomF GmbH, Hamilton Health Sciences, Holcim, IHF, Innosuisse, IumiraDX, Kantar, LabPoint, MedAlliance, Medcon International, Medical Education Global Solutions, Medtronic, MicroPort, Monocle, Novartis, Novo Nordisk, OM Pharma, Pfizer, Quintiles Switzerland Sarl, RecorMedical, Roche Diagnostics, Roche Pharma, Sahajanand IN, Sanofi, Sarstedt AG, Servier, Sorin SRM SAS, SSS Int., Terumo Deutschland, Trama Solutions, V- Wave, Vifor, and ZOLL; these grants do not impact on Dr Ruschitzka’s personal remuneration. Dr Ludewig is also member of the board of Stromal Therapeutics. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.