Ventricular assist device for a coronavirus disease 2019-affected heart

Clinical manifestation and evolution

A 30-year-old male patient with a history of idiopathic cardiomyopathy with reduced LVEF was referred to our emergency department with clinical signs of respiratory failure. He presented with fever, tachypnoea, and laboratory tests indicating sepsis. Initial N-terminal prohormone of brain natriuretic peptide (NT-proBNP) was 71 022 pg/mL, C-reactive protein was 16.1 mg/dL, interleukin 6 (IL-6) was 282 pg/mL, and procalcitonin (PCT) was 0.5 ng/mL (Figure 1 S1). During his repeated outpatient surveillance visits before, he had never presented with signs of decompensation or exertional dyspnoea and always presented with a stable disease with a 3D LVEF of 23%. Endomyocardial biopsy performed 5 months before the acute episode did not show acute inflammatory involvement nor signs for amyloidosis or M. Fabry. Initial real-time qRT-PCR testing as well as further real-time qRT-PCR and IgG-antibody tests for SARS-CoV-2 were positive during the entire hospitalization (Table 1). The chain of infection was not investigated.

Respiratory and low-output heart failure required early oral intubation and implantation of a veno-arterial extracorporeal membrane oxygenation (VA-ECMO) system directly at admission. Echocardiography further confirmed terminal heart failure and reduced contractility [severely impaired systolic LV function (3D LVEF 6%, global longitudinal strain −1.0%), LV dilatation (LV end-diastolic diameter 8.5 cm, E-point septal separation distance 3.6 cm, LV end-diastolic volume 400 mL), impaired right ventricular function (tricuspid annular plane excursion 9 mm), RV dilatation (RV end-diastolic diameter basal 6.4 cm), and severe functional mitral regurgitation]. Later, the partial recovery of respiratory functions and the VA-ECMO allowed for respirator weaning. Owing to prolonged haemodynamic instability and cardiogenic shock, without realistic option for VA-ECMO weaning, the emergency heart failure board decided for cardiac surgery based on the patient's and patient's family consent. Nevertheless, owing to prolonged haemodynamic instability and cardiogenic shock, the emergency heart failure board decided for cardiac surgery based on the patient's consent. At that time, NT-proBNP was 163 837 pg/mL, C-reactive protein raised to 24.5 mg/dL, IL-6 to 124 pg/mL, and PCT to 7.73 ng/mL (Table S1). The patient was referred to heart surgery with LVAD as therapeutic bridge-to-transplant option, percutaneous temporary right ventricular assist device (RVAD) implantation,³ and tricuspid repair on Hospital Day 13. Hereafter, weaning from the RVAD was successful 2 weeks after implantation. Imaging confirmed successful recovery from COVID-19-associated sepsis and cardiogenic shock (Figure 1). The patient was further initially treated with convalescent serum of COVID-19 patients for 5 days. Initial angiotensin-converting enzyme (ACE) inhibitor treatment (ramipril 5 mg twice daily) was discontinued at the time of hospital admission.

Surgical myocardial biopsy

Because SARS-CoV-2 may directly affect pre-diseased hearts,⁴ a heart sample obtained during surgery was used for evaluation. Conventional, immunohistochemical, transmission electron, and light sheet fluorescence microscopy (LSFM) were performed on a heart sample obtained during cardiac surgery. Histology and transmission electron microscopy (TEM) confirmed moderate dilated cardiomyopathy but no lymphocytic myocarditis (Figure 1C and Figure S4). As by real-time qRT-PCR, as reported in other patients with fatal outcomes from COVID-19,^{4, 5} SARS-CoV-2-RNA was not detected within the myocardium, and TEM was used as it might discover coronavirus particles in the hearts of patients.⁶ We found structures with high similarities to SARS virus within the patient's heart cells (Figure 1). We did not find viral factories indicating virus replication, but multivesicular bodies, which are similar in structure (Figure S5). We 3D-analysed expression levels of the ACE2 receptor and transmembrane protease, Serin 2 (TMPRSS2), potentially required for cardial viral uptake.⁷ LSFM⁸ could show focal ACE2 and TMPRSS2 expression (Figure 1). In-depth analysis showed a spatial overlap of ACE2 expression with SARS-CoV-2 spike protein staining performed with specific antibodies, further implicating the presence of SARS-CoV-2 in the patient's heart (Figure 1). Similar findings could be obtained using SARS-CoV-2 nucleocapsid protein staining (Figure S1).