Intramyocardial Inflammation after COVID-19 Vaccination: An Endomyocardial Biopsy-Proven Case Series

Myocarditis in response to COVID-19 vaccination has been reported since early 2021. In particular, young male individuals have been identified to exhibit an increased risk of myocardial inflammation following the administration of mRNA-based vaccines. Even though the first epidemiological analyses and numerous case reports investigated potential relationships, endomyocardial biopsy (EMB)-proven cases are limited. Here, we present a comprehensive histopathological analysis of EMBs from 15 patients with reduced ejection fraction (LVEF = 30 (14–39)%) and the clinical suspicion of myocarditis following vaccination with Comirnaty® (Pfizer-BioNTech) (n = 11), Vaxzevria® (AstraZenica) (n = 2) and Janssen® (Johnson & Johnson) (n = 2). Immunohistochemical EMB analyses reveal myocardial inflammation in 14 of 15 patients, with the histopathological diagnosis of active myocarditis according the Dallas criteria (n = 2), severe giant cell myocarditis (n = 2) and inflammatory cardiomyopathy (n = 10). Importantly, infectious causes have been excluded in all patients. The SARS-CoV-2 spike protein has been detected sparsely on cardiomyocytes of nine patients, and differential analysis of inflammatory markers such as CD4+ and CD8+ T cells suggests that the inflammatory response triggered by the vaccine may be of autoimmunological origin. Although a definitive causal relationship between COVID-19 vaccination and the occurrence of myocardial inflammation cannot be demonstrated in this study, data suggest a temporal connection. The expression of SARS-CoV-2 spike protein within the heart and the dominance of CD4+ lymphocytic infiltrates indicate an autoimmunological response to the vaccination.

Meanwhile, large epidemiological studies reveal that the majority of COVID-19vaccine-related myocarditis occur in young male individuals (median age of 21 years) following the second dose of mRNA based vaccines [9]. Most cases (76.8%) have been observed in response to the Pfizer-BioNTech vaccine (Comirnaty ® ), whereas one quarter (23.2%) received the Moderna vaccine (Spikevax ® ). Despite total case numbers, general incidence of myocarditis was shown to be higher in patients receiving Spikevax ® (1.3 to 1.9 (Spikevax ® ), and 0.26 to 0.57 (Comirnaty ® ) extra cases of myocarditis per 10,000 subjects) [10]. Consequently, numerous national authorities recommended the use of Comirnaty ® instead of Spikevax ® for people aged 30 years and younger.
Signs of myocarditis after COVID-19 vaccination usually develop within 2 weeks [11,12] and clinical symptoms generally ease quickly without impairment of cardiac function [13]. Although COVID-19-vaccine-related myocarditis has an overall marginal incidence and affected patients show a fast recovery in absence of short-term complications, myocarditis is an alarming side effect, which needs to be monitored carefully.
While the majority of published cases are based on cardiac magnetic resonance imaging (cMRI) and laboratory evaluation [2][3][4][5][6][7], endomyocardial biopsy (EMB)-proven cases are limited [8,14,15]. Here, we present a comprehensive analysis of EMBs from 15 patients with the suspicion of myocarditis after vaccination against SARS-CoV-2 and reveal a temporal relation between the vaccination and the occurrence of myocardial inflammation, ranging from mild inflammatory cardiomyopathy to severe active myocarditis and giant cell myocarditis.

Results
Patients' characteristics, clinical picture and laboratory findings at hospital admission are depicted in Table 1. Importantly, the suspicion of myocarditis in response to vaccination against SARS-CoV-2 was expressed in all cases. Median left ventricular ejection fraction (LVEF) was 30 (14-39)%. The analyzed cohort ranges from 18 to 68 years of age and is dominated by male sex (9/15, 60%). All patients were Caucasian. Vaccine type, onset of symptoms, clinical suspicion and clinical course differ a lot in the presented cohort (Table 1).
All patients revealed a sudden onset of severe left ventricular failure (LVEF ≤ 45%) and most had additional symptoms such as dyspnea, chest pain, reduced respiratory rate and diminished exercise capacity (Table 1). Data on cardiac MRI were available for six patients, showing no signs of myocarditis in four of them (patients 2, 4, 5 and 11) and typical signs of active myocarditis in two patients (patients 8 and 9) ( Table 1). Five patients showed severe complications including cardiac decompensation (patients 5 and 8, both DCMi) and fulminant cardiogenic shock (patient 14, GCMC), and two patients (patients 2 and 3, AMC and GCMC, respectively) had to be resuscitated before admission to the intensive care unit (Table 1). Data on laboratory cardiac and inflammatory biomarkers including troponin, brain natriuretic peptide (BNP), creatinine kinase (CK) and C-reactive peptide (CRP) are heterogeneous and were elevated in 12 of 15 patients (Table 1).
Immunohistochemical EMB analysis revealed increased presence of inflammation markers in 14/15 patients ( Figure 1, Table 2). Only 1 patient was found to have consistently low levels of inflammatory cells, compatible with the diagnosis of a DCM (patient 12, Figure 1, Table 2). All other patients with myocardial inflammation showed a diverse pattern of inflammatory markers, ranging from mild to severe degree of inflammation. From 14 patients with proof of inflammation, 9 (64%) showed elevated infiltration of CD3 + T cells, 12 (86%) showed increased levels of CD45R0 + T-memory cells and LFA-1 + lymphocytes, 11 (79%) showed increased numbers of MAC-1 + macrophages and HLA-DR + -presenting cells, and none showed augmented levels of perforin + cytotoxic cells ( Figure 1, Table 2). Cellular adhesion molecules ICAM-1 and VCAM-1 were normal in all patients, with the exception of two patient with AMC (patient 2) and GCMC (patient 14), respectively ( Table 2). Further immunohistochemical analysis revealed positive detection of the SARS-CoV-2 spike protein in cardiac tissue in a couple of patients. In particular, the spike protein was found in sparse cells (cardiomyocytes) in 9 of 15 cases ( Figure 2, Table 2). Moreover, except the cases with active myocarditis (patients 1 and 2), giant cell myocarditis (patients 3 and 14), and one case of DCMi (patient 9), CD4 + -T-cell-to-CD8 + -T-cell ratio was ≥1, suggesting a predominantly autoimmunological origin of the observed inflammation ( Figure 3, Table 2).   vaccines, respectively. Immunohistochemical staining was quantified by digital image analysis and CD4-to-CD8 ratio is depicted for each patient in Table 2. Magnification 400×. Scale bars 20 µm.

Discussion
In the present study, we identified 14/15 (93%) patients with suspected myocarditis after COVID-19 vaccination, with EMB-proven myocardial inflammation. Four of them had acute myocarditis (including active myocarditis and giant cell myocarditis), and 10 were diagnosed for inflammatory cardiomyopathy (DCMi). The absence of inflammation in EMBs of one patient might be explained by a sampling error. The onset of symptoms var-ied from 0 to 56 days, which is within the range of previous epidemiological findings [17]. Even though a causal relationship cannot be made, infectious reasons fro myocarditis had been excluded by molecular diagnostics of the most relevant viral pathogens (including SARS-CoV-2). Although latent viral genomes of B19V were found in 10/15 (67%) cases, active B19V transcription was not confirmed in these patients. B19V is the most frequently detected viral species in the human heart, and its contribution to myocardial inflammation and thus, to long-term patient outcome, is highly dependent on its transcriptional activity [18,19]. Therefore, identified latent B19V infections are unlikely causative of the myocardial inflammation in these patients.
Because viral infections have been ruled out as the cause for myocarditis/myocardial inflammation, autoimmunological mechanisms might be an explanation. Cross-reactivity of spike protein antibodies with myocardial contractile proteins, mRNA immune reactivity and hormonal involvement, have been discussed as potential mechanisms by which COVID-19 mRNA vaccines induce hyperimmunity [20]. In the present cohort, the SARS-CoV-2 spike protein was found to be expressed on cardiomyocytes in 9 of 15 patients. Thus, vaccine-encoded spike protein seems to reach the heart, where it might trigger an inflammatory response, resulting in the development of myocarditis or DCMi. Previous mouse data indicated an intramyocardial expression of spike protein after mRNA vaccination [21]; however, human data are scarce. Interestingly, except for five cases (AMC = 2, GCMC = 2 and DCMi = 1), the number of CD4 + T lymphocytes was either equal or higher than those of CD8 + T cells. In acute myocarditis, a shift of CD4-to-CD8 ratio towards CD8 + T cells is known [22], explaining the elevated CD8 + cells in AMC and GCMC patients. However, as CD4 + T cells are considered to be the major driver of autoimmune myocarditis [20], our data support the idea that vaccine-induced myocardial inflammation is a consequence of excessive CD4 + T-cell infiltration, and thus, a potential driver of autoimmunological myocardial damage. Moreover, the expression of HLA-DR was increased in 11 of 14 (79%) patients with evident inflammation. Due to the fact that HLA class II regions strongly associate with several autoimmune diseases [20], induction of HLA-DR in response to the vaccination supports an autoimmunological contribution to myocardial inflammation after vaccination. Expression of ICAM-1 and VCAM-1 adhesion molecules on the cardiac (micro)vascular endothelium is upregulated upon infection with myocarditis-associated viruses [23]. Besides one patient with AMC (patient 2) and one patient with GCMC (patient 14), none of the patients showed increased levels of ICAM-1 and VCAM-1, supporting a non-infectious origin of the observed inflammation. Perforin-mediated cardiac damage is known to be involved in viral [24,25] and acute idiopathic myocarditis [26]. Interestingly, none of the patients showed increased presence of perforin + cells, indicating no contribution of cytotoxic events following the COVID-19 vaccination. Moreover, data on perforin + cells in patients with reduced LVEF show that the lack of perforin + cardiac infiltrates is associated with spontaneous LVEF improvement [23]. Thus, improvement of cardiac function could be assumed, as described for most of the vaccine-related myocarditis patients [13].
Among EMB diagnostics, non-invasive imaging techniques have been used to identify COVID-19-vaccine-related myocarditis [2][3][4][5][6][7]. Besides cMRI markers such as T2-weighted ratio, early gadolinium enhancement and late gadolinium enhancement within the Lake Louise criteria and somatostatin receptor positron emission tomography/computed tomography (PET/CT) have been applied for the diagnosis of myocarditis following COVID-19 vaccination [27]. However, as cardiac imaging techniques give helpful, but limited information about severe myocardial inflammation, they are unable to correlate with mild inflammation and viral persistence. The present case series reveal that cMRI was able to detect myocardial inflammation in only 33.3% (2/6) of the cases (Table 1). Thus, EMB remains the definite standard in the diagnosis of viral and inflammatory cardiomyopathy.

Study Limitations
There are several limitations of this study which have to be considered when interpreting the obtained results. First, no causality can be assumed or established due to the lack of a control group and the observational character of a case series. Although other causes of myocardial inflammation have been ruled out, there is no direct evidence of a vaccine-induced inflammatory response in the presented cases. Second, an incidence of vaccine-mediated myocarditis/inflammatory cardiomyopathy cannot be determined due to the selective approach of choosing cases with the suspicion of myocarditis after COVID-19 vaccination and the simultaneous availability of EMBs. Third, there is a lack extended clinical data for all of the patients covered in this multicenter study, and clinical data and evaluations were collected from different physicians. Finally, vaccine type and onset of symptoms after first or second dose differs between the cases and are therefore difficult to compare.

Patients and Clinical Investigations
In total, 15 patients (median age 38 (31-52) years; 9 men) from 11 different clinics were included in this this multicenter study. For all patients, the clinical suspicion of myocarditis after COVID-19 vaccination was expressed by the physician in charge due to a temporal relationship of the vaccination and the onset of myocarditis symptoms. All patients underwent key clinical investigations including laboratory testing, echocardiography and electrocardiogram, and some patients (n = 6) underwent additional cardiac magnetic resonance imaging (cMRI) [28]. Furthermore, EMBs were obtained from each patient and sent to the Institute for Cardiac Diagnostics and Therapy (IKDT, Berlin, Germany), for routine workup including histology, immunohistochemistry and molecular virology. The decision to perform EMB was made by the attending physician based on the setting of unexplained heart failure and sudden onset of severe left ventricular failure [29]. Coronary artery disease and other possible causes of myocardial dysfunction have been excluded by angiography prior to EMB in all patients. Left ventricular ejection fraction (LVEF) was determined by echocardiography with the Simpson method. All numerical data are presented with median and interquartile range.

Conclusions
The present study summarizes EMB-based diagnostics of 15 patients with clinical suspicion of myocarditis following vaccination against SARS-CoV-2. It identifies 14 of 15 patients with myocardial inflammation, ranging from inflammatory cardiomyopathy to active myocarditis and severe giant cell myocarditis. Although a causal relationship between vaccination and the occurrence of myocardial inflammation cannot be established based on the findings, the cardiac detection of spike protein, the CD4 + T-cell-dominated inflammation and the close temporal relationship argue for a vaccine-triggered autoimmune reaction.