Pericarditis and myocarditis long after SARS-CoV-2 infection: a cross-sectional descriptive study in health-care workers

Evidence before this study Very little evidence exists describing long cardiac sequelae after SARS-CoV-2 infection. Although pericarditis and myocarditis are the two most frequent cardiac manifestations observed after a viral infection, as of May 13, 2020, the peer-reviewed literature was limited to isolated case reports of myocarditis and pericarditis during the COVID-19 hospitalization phase and to a retrospective observation in 26 recovered patients with COVID-19 pneumonia presenting cardiac complaints during hospitalization, revealing the presence of myocardial oedema in 14 (54%) patients and late gadolinium enhancement in 8 (31%) patients. These small size case series, limited to hospitalized RT-PCR patients with COVID-19 pneumonia, are insufficient to generalize conclusions about the true prevalence of pericardial and myocardial long involvement after SARS-CoV-2 infection. In addition, no study has investigated the immunological consequences of SARS-CoV-2 infection in the settings of pericarditis and myocarditis.

Added value of this study To our knowledge, this is the largest cohort of subjects (N=139) —even for other common viruses— with clinical, electrocardiographic, laboratory and CMR imaging evaluations, to assess pericardial and myocardial involvements after SARS-CoV-2 infection. The strength of this study is the addition of non-hospitalized participants and also the inclusion of participants diagnosed of past SARS-CoV-2 infection through serology. Contrary to previous studies, women are well represented. We found a prevalence of pericarditis or myocarditis up to 40% cases; pericarditis coexisted with some degree of concurrent myocardial inflammation in 11% cases. Study participants who were previously hospitalized for COVID-19 pneumonia and patients who received antiviral (hydroxychloroquine, lopinavir-ritonavir) or anti-inflammatory (high-dose glucocorticoids and anti-interleukin treatments) treatments, and who were on chronic drug treatment with statins, were less likely to develop pericarditis or myocarditis. The clinical assessment of the participants showed clinical stability without any patient presenting severe pericardial effusion, heart failure or left ventricular dysfunction. We provide new data on seropositive subjects; although RT-PCR participants were more likely to still present symptoms than participants diagnosed by serology, the prevalence of pericarditis, myocarditis or myocarditis, almost three months after the initial viral prodrome, was high in both groups. In-depth investigation of the distribution of multiple major and minor populations of immune cells in blood showed high frequency of altered immune profiles after SARS-CoV-2 infection. The altered immune cell profiles identified partially mimic abnormalities previously reported during active infection together with others described here for the first time, with unique patterns associated with pericardial and/or myocardial injury. Nonetheless, we also described altered immune profiles in participants without pericardial and myocardial manifestations. Whether these later alterations are due to persistence of tissue damage in other organs affected by SARS-CoV-2, such as the lung, or they reflect normal post-infection immune recovery mechanisms, remains to be investigated.

Implications of all the available evidence At present, there is much interest in the long-term sequelae of COVID-19. It is intriguing that pericarditis and myocarditis were observed so long after SARS-CoV-2 infection and also in some presently asymptomatic subjects, in association with notably altered immune cell profiles in blood. These observations will probably apply to the general population infected and may indicate that cardiac sequelae might occur late, paving the way for a better understanding the immune mechanisms involved. Thus, our study may have health-care consequences given the widespread diagnosis of SARS-CoV-2 infection in population-based seroprevalence studies.