Cardiac MRI Findings in Patients Clinically Referred for Evaluation of Post-Acute Sequelae of SARS-CoV-2 Infection

Persistent or recurrent cardiovascular symptoms have been identified as one of the hallmarks of long-COVID or post-acute sequelae of SARS-CoV-2 infection (PASC). The purpose of this study was to determine the prevalence and extent of cardiac abnormalities in patients referred for cardiac MRI due to clinical evidence of PASC. To investigate this, two tertiary care hospitals identified all patients who were referred for cardiac MRI under the suspicion of PASC in a 2-year period and retrospectively included them in this study. Patients with previously known cardiac diseases were excluded. This resulted in a total cohort of 129 patients (63, 51% female; age 41 ± 16 years). The majority of patients (57%) showed normal cardiac results. No patient had active myocarditis or an acute myocardial infarction. However, 30% of patients had evidence of non-ischemic myocardial fibrosis, which exceeds the prevalence in the normal adult population and suggests that a possible history of myocarditis might explain persistent symptoms in the PASC setting.


Introduction
The COVID-19 pandemic caused by the Severe Acute Respiratory Syndrome Coronavirus-2 (SARS-CoV-2) has affected millions of individuals worldwide. Although COVID-19 is primarily a respiratory illness, it has become evident that it can also lead to cardiovascular complications. While the peak of the pandemic seems to have subsided, recent studies have highlighted potential long-term cardiovascular complications following the acute COVID-19 illness.
In fact, persistent or recurrent cardiovascular symptoms after recovery from the acute phase of COVID-19 have been identified as one of the hallmarks of post-acute sequelae of SARS-CoV-2 infection (PASC) [1]. Typical cardiovascular symptoms include fatigue, exertion dyspnea, exercise intolerance, tachycardia/palpitations, and chest pain [2]. While the underlying pathophysiology for these symptoms is not entirely understood and is the subject of ongoing research, it is believed to be multifactorial, including direct viral damage and immune dysregulation, potentially leading to persistent inflammation [3].
Current reports on the prevalence of cardiovascular PASC vary widely [1,3]. Possible explanations for this include inconsistent inclusion criteria and varying definitions of acute myocardial inflammation [1,3,4]. In an effort to harmonize inclusion criteria for studies on this topic, a recent consensus statement by the American College of Cardiology concluded that unselective screening of low-risk individuals is not warranted, but in the presence of persistent cardiorespiratory symptoms, cardiac MRI should be clinically considered [4]. To further specify this, the Society for Cardiovascular Magnetic Resonance has proposed optimized cardiac MRI protocols for scanning patients with active or convalescent phase COVID-19 infection and stresses that the evaluation of imaging should be strictly based on established and validated criteria such as the 2018 Lake-Louise criteria [5].
The scope of this two-center retrospective study was therefore to determine the presence and extent of cardiac abnormalities in patients with signs of PASC and clinical referrals for cardiac MRI using the previously proposed protocols and evaluation techniques.

Study Population
The institutional ethics commissions of both participating centers approved this retrospective study with a waiver for informed consent. Both participating tertiary care hospitals retrospectively identified all patients who were referred for cardiac MRI within a 2-year period (May 2020 to May 2022) due to clinical signs of PASC according to current recommendations (fatigue, exertional dyspnea, exercise intolerance, tachycardia/palpitations, and chest pain > 4 weeks after infection [1,2]). Patients with previously known cardiac disease were excluded from the study.

Cardiac MRI
Institutional imaging protocols were in accordance with recommendations for scanning COVID-19 patients in the convalescent phase [5]. They included a short-axis cine stack covering the entire LV-extend, at least two long-axis cine views, parametric T1 and T2 mapping, as well as the administration of an intravenous contrast agent with subsequent acquisition of late gadolinium enhancement (LGE) sequences in both short-and long-axis views.

Image Analysis
Cardiovascular radiologists with at least 5 years of experience (M.C.H., A.I.) performed consensual image analysis using dedicated post-processing software, including assessment of myocardial edema and LGE. Left-ventricular ejection fractions and end-diastolic volume indices were assessed in a qualitative manner using cut-offs of <55% for the ejection fraction and ≥100 mL/m 2 for the end-diastolic volume index. Pericardial and pleural effusions were measured on axial slices and reported if they exceeded a thickness of 5 mm (pericardial effusion) and 20 mm (pleural effusion), respectively. For the evaluation of parametric mapping, a global measurement approach was used, and relaxation times of >2 standard deviations from the respective local reference values were considered abnormal to accommodate mapping sensitivity due to the differing field strengths used in this bicentric design (site 1: 1.5 Tesla, site 2: 3 Tesla) [6]. The presence of LGE was judged on a qualitative basis and, if present, further specified as ischemic or non-ischemic patterns based on localization and distribution.

Statistical Analysis
Continuous data are expressed as the mean ± standard deviation. Categorical variables are expressed as frequencies, with their respective proportions expressed in percentage.

Discussion
In this pre-selected cohort of patients referred for cardiac MRI with clinical suspicion of PASC, 30% of patients had signs of post-inflammatory myocardial fibrosis, which might indicate a possible history of myocarditis [4]. In 4% of patients, post-ischemic myocardial findings were found. No patient had active myocarditis or an acute myocardial infarction. Previously unknown structural cardiac disease was diagnosed in 9% of patients. In over half (57%) of the patients, cardiac results were normal; in 5% of these patients, pulmonary abnormalities were suspected to be a possible explanation for persistent symptoms.
Previous studies have reported a wide range of prevalence for abnormal cardiac MRI findings in patients after SARS-CoV-2 infection depending on study design and inclusion criteria (26-78%, [3,7]). To improve generalizability and more closely represent a real-world scenario in imaging departments when trying to assess PASC-associated cardiovascular disease, this study focused solely on clinical referrals for cardiac MRI and strictly relied on established international guidelines for the detection of myocardial inflammation [4]. In this study, 30% of patients had non-ischemic myocardial fibrosis, which exceeds the prevalence in the normal adult population (general rates of approximately 7.9% of non-ischemic scars without direct clinical consequence have been reported in the US [8]). In a recent study, serial cardiac MRI investigations in both symptomatic and asymptomatic individuals (baseline at >4 weeks after SARS-CoV-2 infection, follow-up within a year) revealed that the presence of non-ischemic LGE at baseline was associated with the persistence or new onset of symptoms at follow-up [7]. However, in contrast to previous clinical studies, which showed rates of 8-12%, this study did not identify cases of active myocardial inflammation [3]. This is in line with results from a meta-analysis of autopsy results in which histopathological evidence of active myocarditis was rarely (<2%) found [9]. A potential explanation could be the methodological differences between previous clinical studies and the current one, namely the different time intervals between SARS-CoV-2 infection and cardiac MRI. While a longer time difference could have introduced bias through normalization of T2-relaxation times over time, it is also important to acknowledge that shorter time differences, especially those falling below the PASC cut-off of 4 weeks after infection, might represent reverberations of the initial infection [3].
However, due to the limitations of the retrospective design and the lack of baseline studies before COVID-19, there is no proof of a causal relationship between SARS-CoV-2 infection and structural, ischemic, or post-inflammatory findings. Especially for findings such as dilated cardiomyopathy, these could also be perceived as previously subclinical diseases that became overt in the PASC setting. Another limitation of this study that merits consideration is the absence of blind inter-and intra-reader variability assessments. However, since this study focused on the qualitative assessment of the cardiac abnormalities and only a few quantitative measurements were assessed, additional quantitative measurements would be of limited incremental value.

Conclusions
In this two-center cardiac MRI study, the majority of patients (57%) with clinical suspicion of cardiovascular PASC showed normal cardiac results. However, 30% of patients had evidence of non-ischemic myocardial fibrosis (exceeding the prevalence in the normal adult population), suggesting that a possible history of myocarditis might be an explanation for persistent symptoms in a PASC setting. Further previously unknown cardiac abnormalities like post-ischemic fibrosis (4%) and structural heart disease (9%) were also found as possible correlates for persistent clinical symptoms, although a causal relationship to COVID-19 seems unlikely. Data Availability Statement: The datasets analyzed during the current study are available from the corresponding author on reasonable request.