Cardiac Injury Before and After COVID-19

Recent studies based on late gadolinium enhancement (LGE) cardiac magnetic resonance (CMR) have raised great concern about frequent residual myocardial injury even after mild or asymptomatic COVID-19.^{1 , 2} However, although signs of myocardial injury were found in large proportions of patients after COVID-19, all studies published to date lack baseline imaging and are therefore unable to discriminate between preexisting and COVID-19-induced injury. Against this background, we performed a longitudinal study to assess the individual cardiac impact of COVID-19.

A prospective registry of patients undergoing LGE CMR in the context of atrial fibrillation was screened for patients with documented SARS-CoV-2 infection subsequent to LGE CMR. Eligible patients then underwent post-COVID-19 LGE CMR using the same scanner and sequence as for pre-COVID-19 LGE CMR. T1-weighted inversion recovery gradient-echo sequences were acquired in sinus rhythm using electrocardiographic gating and a free-breathing 3-dimensional navigator 15 to 20 minutes after administering an intravenous bolus of 0.2 mmol/kg gadobutrol. An inversion time scout sequence was used to determine the optimal inversion time that nulled the left ventricular myocardial signal. LGE was independently assessed qualitatively by 2 experienced investigators blinded to patient information. For quantitative analyses, a 3-dimensional reconstruction of the left ventricle was performed using dedicated software (Adas3D Medical). LGE was then automatically quantified on the basis of a prespecified signal intensity threshold of ≥3 SDs above the mean of a remote nonenhanced myocardial region. Approval was obtained from the local research ethics committee, and written informed consent was obtained from each patient.

Thirty-one patients with confirmed COVID-19 between March 2020 and February 2021 were included. Seven patients (23%) had been hospitalized at the time of acute presentation with COVID-19, of whom 2 (6%) required intensive care. Most patients (29 [96%]) had been symptomatic, but none reported cardiac symptoms.

At a median of 5 months post-COVID-19, LGE lesions indicative of residual myocardial injury were encountered in 15 of the 31 patients (48%; not considering LGE lesions at the right ventricular insertion points), which is in line with previous reports.^{1 , 2} The majority of lesions were located midmyocardially (65%), with only a few isolated subendocardial (18%) or transmural (18%) lesions. However, intraindividual comparison with the pre-COVID-19 CMR revealed all of these lesions as pre-existing with identical localization, pattern, and transmural distribution and thus not COVID-19-related (Figure 1 ). Quantitative analyses, performed independently, detected no increase in the size of individual LGE lesions nor in the global left ventricular LGE extent. Comparison of pre- and post-COVID-19 cine imaging sequences did not show any differences in ventricular functional or structural parameters.

Figure 1.

Corresponding LGE Lesions Pre- and Post-COVID-19

(A to C) Representative examples of corresponding lesions in pre- and post-COVID-19 scans from 3 different patients. (Left) 3-dimensional reconstructions of the left ventricle with late gadolinium enhancement (LGE)–based color coding. (Middle and right) T1-weighted short-axis slices with and without LGE color coding of the respective layer. Arrows indicate LGE.

Of course, these findings are not generalizable and in particular may not apply to patients with severe COVID-19 and/or evident cardiac involvement. Unfortunately, although none of the patients displayed elevated troponin or C-reactive protein levels at the time of post-COVID-19 CMR, troponin levels during acute COVID-19 as a surrogate for cardiac involvement were not acquired systematically, because patients were included only after recovery. Moreover, although T1/T2 mapping would allow for better discrimination of ongoing inflammatory processes and edema from chronic scarring, such CMR sequences were not performed in the pre-COVID-19 scans and were therefore not available for intraindividual comparisons.

Because patients were recruited from our prospective atrial fibrillation registry, all patients had a history of atrial fibrillation, and consequently a large proportion displayed cardiovascular risk factors such as hypertension (58%), dyslipidemia (45%), diabetes (26%), coronary artery disease (10%), and history of heart failure (6%). On the basis of available data linking atrial fibrillation and cardiovascular risk factors to COVID-19-associated complications, this selected cohort would be considered at particularly high risk not only for preexisting cardiac injury but also for cardiac sequelae of COVID-19. Against this background, the complete absence of de novo LGE lesions is specifically noteworthy.

To the best of our knowledge this is the first CMR study to assess myocardial injury pre- and post-COVID-19. Although with only 31 patients longitudinally studied we cannot rule out the possibility of rare events of COVID-19-induced myocardial injury; the complete absence of de novo LGE lesions after COVID-19 in this cohort indicates that outside special circumstances, COVID-19-induced myocardial injury may be much less common than suggested by previous studies.