By Michael H. Crawford, MD
Professor of Medicine, Lucy Stern Chair in Cardiology, University of California, San Francisco
SYNOPSIS: In a review of COVID-19-associated acute myocarditis, investigators learned it is a rare complication that can occur without concomitant pneumonia, and frequently presents as cardiogenic shock. With supportive therapy, the short-term mortality rate is low.
SOURCE: Ammirati E, Lupi L, Palazzini M, et al. Prevalence, characteristics, and outcomes of COVID-19-associated acute myocarditis. Circulation 2022;145:1123-1139.
Although COVID-19 illness can affect the heart in many ways, acute myocarditis (AM) is thought to be a rare complication. Little is known about its prevalence, characteristics, or outcomes. Ammirati et al conducted an international, multicenter, retrospective, observational study in 23 cardiology centers to answer these questions.
Patients with documented COVID-19 infection, symptoms consistent with AM, and who met definite criteria for myocarditis or probable myocarditis were considered for the analysis. Definite AM criteria included endomyocardial biopsy or autopsy-positive, or by elevated cardiac biomarkers and cardiac magnetic resonance imaging (MRI) findings consistent with myocarditis. Probable AM criteria were positive biomarkers and electrocardiogram (ECG) or echocardiographic evidence of AM with a follow-up MRI (within nine months) consistent with non-ischemic, late gadolinium enhancement.
The authors excluded patients with a previous diagnosis of ischemic heart disease (IHD) or cardiomyopathy, patients younger than age 10 years, men older than age 50 years, women older than age 55 years without a negative test for IHD, and patients older than age 70 years without a positive biopsy. Researchers also assembled a comparator group of AM patients from the pre-COVID-19 Lombardy registry with similar inclusion criteria.
Clinical data collected included fulminant AM, defined as fast evolution with hemodynamic compromise and shock, which was categorized at each site. Among the 56,963 hospitalized patients with COVID-19 infection, after applying the exclusion criteria, there were 54 patients with definite or probable AM, for a mean incidence of 0.0024 or 2.4 cases/1,000 hospitalized patients. Since not all suspected AM patients had undergone the appropriate tests to establish the diagnosis, there were 112 cases of possible AM, for an incidence of 4.1 cases/1,000 hospitalized patients. The median age of the 54 definite/probable cases was 38 years; 39% were women, and 77% were non-Hispanic white. None of the patients were vaccinated before the onset of COVID-19 infection.
On admission, chest pain (56%) and dyspnea (54%) were the most common symptoms. In 57%, AM occurred in the absence of pneumonia. Those with pneumonia were older than those without (mean age = 45 years vs. 30 years; P = 0.002). A fulminant presentation requiring inotropic or temporary circulatory support occurred in 39%. The hospital mortality rate was 5.5%, and the 120-day mortality rate was 6.6%, but 15% in those with pneumonia vs. 0% in those without (P = 0.04). Among the 19% who required mechanical circulatory support, the 120-day mortality rate was 38%. Most patients received corticosteroids (56%) and a minority received other anti-inflammatory or anti-viral therapy. In the survivors (n = 47), the mean ejection fraction (EF) on admission was 40%, and was 55% on discharge (P < 0.0001).
The authors concluded AM is rare in hospitalized patients with COVID-19 and often occurs without associated pneumonia. Hemodynamic instability is common, and the 120-day mortality rate is about 7%, occurring exclusively in those with associated pneumonia.
COMMENTARY
Most papers on COVID-19 and heart-related side effects have not been particularly helpful, but this large database study seemed to include clinically useful information. The authors reported AM occurs in two to four times per 1,000 hospitalized COVID-19 patients with and without pneumonia. Cardiogenic shock occurs in almost 40%; pneumonia exacerbates the prognosis. All of the 6% who died in a hospital had pneumonia. Corticosteroids were used commonly and appeared to be safe. EF improved in survivors, and short-term follow-up displayed excellent recovery. Ammirati et al also compared their data to the Lombardy AM database of 443 patients from the pre-COVID era with similar inclusion criteria. This revealed fulminant AM was more common with COVID-19 AM (39% vs. 8%) and that initial EF was lower (40% vs. 55%). Not surprisingly, the short-term mortality rate was higher as well (7% vs. 1%). Thus, COVID-19 AM seems to be a more severe disease than prior viral AM cases. Of interest, no one with COVID-19 AM had been vaccinated.
The relatively large population of COVID-19 patients (56,963) in whom AM was diagnosed using strict criteria involving biopsies and MRI is commendable. However, this was an observational study, so selection biases may have occurred. Also, the lack of MRI data for all patients probably kept the prevalence of AM lower. The authors excluded 43 cases because there were insufficient data to support the diagnosis of AM (but AM still was suspected clinically).
Ammirati et al performed some complex estimating to account for these issues, believing the incidence could be as high as 0.57 compared to the 0.0024-0.0041 they reported. In addition, there was no central review of the biopsy or MRI data, although standard criteria were used. Finally, there was no systematic screening for other viruses that could have caused AM, but parvovirus B19 was discovered in one-third of the biopsies. Parvovirus B19 likely is an innocent bystander because it can be seen in normal biopsies, but it could play a role here. Regarding the etiology of COVID-19 AM, viral genome was found in only one-quarter of patients biopsied, which would favor an immune response mechanism for AM in most rather than direct viral effects.
