CMR in the Era of COVID-19

Corresponding Author

The coronavirus disease-2019 (COVID-19) pandemic continues to expand in its global impact: it has infected nearly 5 million people and has been responsible for >300,000 deaths worldwide as of May 20, 2020 (according to the Johns Hopkins Corona Virus Resource Center) (1). Although the most profound manifestations of this disease have been due to pulmonary complications, our understanding of the devastating cardiac complications of this disease has been growing. In early studies from Wuhan, ∼7% of hospitalized patients with COVID-19 had evidence of increased troponin levels, indicating myocardial involvement, which was associated with worse outcomes (2). COVID-19 has been associated with heart failure, acute coronary syndromes, myocarditis, and right heart dysfunction. The exact mechanism of cardiac injury in COVID-19 has not been well established but may involve direct myocyte involvement, or secondary injury from the profound inflammatory response invoked by the virus (3). With the growing number of infections, we will likely continue to see the impacts of persistent cardiac symptoms in patients infected with COVID-19 for years to come.

Given the typical manifestations of cardiac involvement in COVID-19, namely myocarditis, right heart failure, and acute coronary syndromes, cardiac magnetic resonance (CMR) would theoretically be the ideal technique for assessing patients with COVID-19 and suspected cardiac involvement. CMR is the most accurate cardiovascular imaging technique for diagnosing myocarditis and for differentiating this from acute myocardial infarction or other etiologies such as stress-induced cardiomyopathy (4). Furthermore, CMR is a reference standard technique for quantifying right heart dysfunction. However, in the COVID-19 era, performance of CMR in patients in hospitalized patients with active symptoms has to be weighed against the potential of the exposure of staff and the contamination of scanners, and the likelihood that the information derived from CMR will have an immediate implication for patient management to justify risks. For these reasons, the use of CMR in the acute phase of COVID-19 infections has been limited. As progressively more patients are recovering from the acute phase of this disease, continued manifestations such as chest pain, heart failure, and palpitations are increasingly recognized as ongoing sequelae in the convalescent phase of the disease. To date, there has been limited evaluation of the CMR imaging findings in patients with ongoing symptoms who have recovered from the acute disease. Questions regarding the severity of myocardial scarring and ongoing inflammation in patients with persistent cardiac symptoms have not been fully explored, with only a few case reports presented.

In this issue of iJACC, Huang et al. (5) describe their initial findings in a retrospective single-center observational study of 26 patients who were diagnosed with COVID-19, had confirmation of recovery following a 14-day isolation period and 2 negative COVID-19 test results, and were subsequently referred to undergo CMR for assessment of chest discomfort or palpitations. None of these patients had any significant cardiac history or coronary risk factors except a history of hypertension in 2 individuals (8%). A comprehensive CMR evaluation was performed on a 3-T scanner including conventional techniques of cine, late gadolinium enhanced imaging (LGE), and T2-weighted short-tau inversion recovery (STIR) imaging, as well as parametric mapping of myocardial T1, T2, and extracellular volume (ECV).

Twenty-two (85%) of the patients had moderate COVID-19 pneumonia, and the other 4 had severe pneumonia (4). None of these patients appeared to have developed symptoms or signs of cardiac failure during their initial treatment for COVID-19 pneumonia. The median time between onset of cardiac symptoms and CMR was 47 days. Only 13 of the 26 patients had high-sensitivity cardiac troponin I (hs-cTnI) levels measured during their COVID-19 hospitalization, and all had a normal hs-cTnI level at the time of CMR. Fourteen (54%) patients had evidence of increased T2 signal ratio according to STIR imaging, and 31% of cases had evidence of small focal subepicardial and patchy mid-wall LGE in a pattern consistent with myocarditis. Abnormal T2 STIR and LGE were noted to involve 33% and 4% of segments, respectively, suggesting significant myocardial edema with minimal scarring. Only 1 patient, who had positive conventional CMR findings based on T2 STIR and LGE, had an abnormal ejection fraction with evidence of a regional wall motion abnormality in the segments with myocardial edema. Although the mean left ventricular function was similar among control subjects and subjects with and without conventional CMR abnormalities, the cohort with conventional CMR findings had reduced right ventricular function compared with healthy control subjects (5). When divided into 2 groups based on the presence of conventional CMR findings, the group with positive CMR findings had increases in global T1, T2, and ECV compared with those without conventional CMR findings and normal control subjects. Notably, in the 8 LGE-positive patients, the remote myocardium had evidence of increased T1, T2, and ECV compared with healthy control subjects. These findings suggest the presence of mild edema based on the modest magnitude of increase in T1, T2, and ECV, respectively (1,271 vs. 1,224 ms, 43 vs. 39 ms, and 28.2% vs. 23.7%).

The current study (5) is one of the first and largest case series showing that CMR detects abnormal findings in patients who have recovered from COVID-19 who had no evidence of cardiac involvement during their initial treatments but present subsequently with nonspecific cardiac symptoms. The findings showed globally increased T1, T2, and ECV in patients with conventional CMR abnormalities, even in remote segments, which raise questions regarding ongoing cardiac inflammation and/or fibrosis. Specifically, there seems to be evidence of diffuse myocardial edema based on T2 STIR imaging disproportionate to the small LGE extent, in an absence of left ventricular dysfunction; this scenario begs the question of whether there is direct viral infection or whether cardiac involvement is primarily due to inflammation. These intriguing questions will need to be studied further in larger studies.

The current study has a number of limitations (4). The retrospective design of the study does not enable assessment of the prevalence of CMR findings in patients who have recovered from acute COVID-19 infections. Second, the cohort of patients is small, although this likely reflects the reduction of advanced cardiovascular imaging utilization and resources in a region within the early parts of the pandemic. Larger studies looking at persistent cardiac involvement in COVID-19 will likely follow as advanced cardiovascular imaging services resume more broadly. Interestingly, the patients had evidence of myocardial edema despite absence of high-sensitive cardiac troponin I positivity, and unfortunately only one-half of the patients had troponin assessments during their hospitalization.

The strong inflammatory response of COVID-19 could potentially provide therapeutic targets for COVID-19 myocarditis, which could necessitate the assessment of COVID-19 myocardial involvement in patients in the acute phase of the disease. There has been particular interest in interleukin-6, and inhibitors of inflammation could potentially be therapies for COVID-19 myocarditis. Because multiple pathologies such as acute myocardial infarction, Takotsubo cardiomyopathy, and myocarditis are all within the differential diagnosis of a positive troponin, chest pain, and heart failure in patients with COVID-19, a rapid CMR protocol could potentially be used to differentiate between these etiologies. However, performing CMR in these patients necessitates the balance between patient benefit and the need to protect staff and critical resources.

The current study by Huang et al. (5) has important implications regarding the study of COVID-19–related myocarditis in showing that CMR abnormalities are present and that active inflammation/edema may extend into the subacute phase. This may enable the study of myocardial involvement in patients at a time when the patients have recovered and are at lower risk for spreading the disease. Furthermore, determining the etiology for increased troponin levels or cardiovascular symptoms could have implications regarding longer term treatment of patients by differentiating evidence of myocardial infarction from myocarditis.

Author Relationship With Industry

Dr. Salerno has received grant support from the National Institutes of Health (R01 HL131919); and research support from Siemens Healthcare. Dr. Kwong has received grant support from the American Heart Association, MyoKardia Inc., and Alnylam Pharmaceuticals.