Abstract Cases of myocarditis, diagnosed clinically by laboratory tests and imaging have been described in the context of mRNA-based anti-SARS-CoV-2 vaccination. Autopsy-based description of detailed histological features of vaccine-induced myocarditis is lacking. We describe the autopsy fndings and common characteristics of myocarditis in untreated persons who received anti-SARS-CoV-2 vaccination. Standardized autopsies were performed on 25 persons who had died unexpectedly and within 20 days after anti-SARS-CoV-2 vaccination. In four patients who received a mRNA vaccination, we identifed acute (epi-) myocarditis without detection of another signifcant disease or health constellation that may have caused an unexpected death. Histology showed patchy interstitial myocardial T-lymphocytic infltration, predominantly of the CD4 positive subset, associated with mild myocyte damage. Overall, autopsy fndings indicated death due to acute arrhythmogenic cardiac failure. Thus, myocarditis can be a potentially lethal complication following mRNA-based anti-SARS-CoV-2 vaccination. Our fndings may aid in adequately diagnosing unclear cases after vaccination and in establishing a timely diagnosis in vivo, thus, providing the framework for adequate monitoring and early treatment of severe clinical cases.
Results Among the 35 cases of the University of Heidelberg, autopsies revealed other causes of death (due to pre-existing illnesses) in 10 patients (Supplementary Table 1). Hence, these were excluded from further analysis. Cardiac autopsy fndings consistent with (epi-)myocarditis were found in fve cases of the remaining 25 bodies found unexpectedly dead at home within 20 days following SARS-CoV-2 vaccination. Main characteristics of the fve cases are presented in Table 2, while further autopsy fndings are shown in Supplementary Table 2. Three of the deceased persons were women, two men. Median age at death was 58 years (range 46–75 years). Four persons died after the frst vaccine jab, the remaining case after the second dose. All persons died within the frst week following vaccination (mean 2.5 days, median 2 days). Clinical fndings, blood tests, ECGs or imaging data were not available as deceased persons did not seek medical attention prior to death. Person 1 was found dead 12 h after the vaccination. A witness described a rat - tling breath shortly before discovering circulatory failure. Person 2 complained about nausea and was found dead soon thereafter. Resuscitation was started immediately but without success, respectively. The other persons were found dead at home without available information about terminal symptoms. According to the available information provided at the time of autopsies, none of the deceased persons had SARS-CoV-2 infection prior to vaccination and nasopharyn - geal swabs were negative in all cases. Histological examination showed infammatory infltra - tion of the myocardium. The infltrate was focal and interstitial in all cases. It was predominantly detected in sections taken from the right ventricular wall and interventricular septum. The histological and immunohistochemical char - acterization revealed that the infammatory infltrate was predominantly composed of lymphocytes. The number of CD3-positive T-cells by far outnumbered the few CD20-pos - itive B-cells detected. In addition, most T-cells belonged to the CD4-positive subset, while only scattered CD8-positive T-cells were seen (Fig. 1, 2, Supplementary Fig. 1/2). The T cells were negative for Tbet as a marker for Th1 cells, GATA3 as a marker for Th2 cells, D2-40, as a marker for Th17 cells (Supplementary Fig. 2). In addition, FOXP3 posi - tive regulatory T cells and CD21 positive follicular dendritic cells were not detected within the cardiac infltrates, while control cases, including cases of sarcoidosis were positive (Supplementary Fig. 2/3). Immunohistochemistry for CD68 showed few interspersed histiocytes (Fig. 2, Supplementary Fig. 1). Microfocal myocyte injury was demonstrable in three cases (patient 1, 2 and 3). No granulomas were found. All cases lacked signifcant coronary heart disease, acute or chronic manifestations of ischaemic heart disease, manifes - tations of cardiomyopathy or other signs of a pre-existing, clinically relevant heart disease. In most cases, an infammatory infltration of the epi - cardium and the subepicardial fat tissue was concomitantly found (cases 2, 3, 4 and 5; Supplementary Fig. 4) and revealed an identical immunophenotype. (T-cell dominant; CD4 > >CD8). In case 2, a prominent CD4-positive lym - phocytic infltration was also recorded at the jab site of the deltoidal muscle (Fig. 3). Analysis for potential infectious agents causing a myocarditis revealed low viral copy num - bers of human herpes virus 6 (HHV6) in one case (case 5). The results of the other four cases were negative for all infectious agents tested, but demonstrated regular amplifca - tion of the GAPDH control suggesting adequate nucleic acid Table quality for analysis.
In three cases, the overall autopsy fndings, in particular presence of (epi-)myocarditis in combination with the absence of other plausible causes of death (especially pulmonary embolism, myocardial infarction, severe brain infarction or bleeding, other cardiac disease), together with the close temporal association with the vaccination event lead to the conclusion that vaccination was the likely cause of (epi-)myocarditis and that this cardiac afection was the cause of sudden death. For case 5, myocarditis was considered to be the cause of death as well, but the detection of HHV6, even in low viral copy numbers provided an alternative explanation for the presence of myocarditis. With regard to the question of a fatal AEFI, case 5 was therefore classifed as “possible”. For case 3 no other cause for the infammatory infltration was found, but the infltrate was discrete and mainly observed in the pericardial fat. Thus, case 3 was categorized as possible AEFI as well. We did not fnd an obvious association between the infltrates and endothelial cells (CD31, D2-40), mesothelial cells (calretinin), or neural cells (S100). During the last 20 years of autopsy service at Heidelberg University Hospital we did not observe comparable myocardial infammatory infltration. This was validated by histological re-evaluation of age- and sex-matched cohorts from three independent periods, which did not reveal a single case showing a comparable cardiac pathology.