Two causes of COVID-19-related myocardial injury-associated cardiogenic shock: Myocarditis and microvascular thrombosis

The diagnosis of COVID-19-related myocarditis often relies on the patient's clinical findings, while pathological evaluation remains challenging.¹ An autopsy study of COVID-19-related myocardial injury reported that only 2%–30% of cases showed clear pathological evidence of lymphocytic myocarditis.^2-4 Notably, some cases showed capillary microvascular thrombosis (MVT) without typical pathological features of myocarditis. Moreover, MVT has been identified as common^{3, 4} and a primary contributor to myocardial injury in COVID-19,⁵ indicating a potential risk of triggering fatal cardiac dysfunction. However, differences in clinical characteristics of the two conditions remain to be evaluated. Here, we report a case involving pathologically diagnosed COVID-19-related lymphocytic fulminant myocarditis (FM) and another case exhibiting clinical features that closely resembled FM and that were associated with COVID-19-related MVT. We aimed to highlight and compare the distinctive characteristics of these cases, shedding light on fatal myocardial injury attributed to MVT in the context of COVID-19.

The patient was a 27 year-old woman with a history of bronchial asthma. She was not vaccinated against COVID-19. A few weeks before, her family member had tested positive for COVID-19 (Table 1). Two days before, she developed chest pain, with worsening symptoms leading to increased fatigue, and visited our hospital. She was in a state of cardiogenic shock. A 12-lead electrocardiogram (ECG) revealed ST-segment elevation (STE) in leads III and V_1–4 (Figure 1). Transthoracic echocardiography (TTE) demonstrated diffuse left ventricular (LV) dysfunction with a left ventricular ejection fraction (LVEF) of 30% (Videos S1 and S2). Laboratory findings revealed elevated levels of cardiac enzymes (Table 2). A COVID-19 antigen test was positive. Coronary angiography (CAG) showed no significant stenosis. Right ventricular (RV) endomyocardial biopsy (EMB) was performed. The patient's condition deteriorated, leading to the initiation of mechanical circulatory support (MCS) [veno-arterial extracorporeal membrane oxygenation (VA-ECMO) and intra-aortic balloon pump] (Figure 2A). Methylprednisolone (1000 mg for 3 days), along with remdesivir (200 mg for 10 days) and sotrovimab (single dose, 500 mg) was started. On day 2, LV and RV function further declined. Cardiac function began to improve and had nearly normalized by day 5, allowing for weaning from MCS. The patient was discharged on day 36 without any symptoms.

EMB revealed diffuse lymphocytic infiltration and myocardial cell injury, which are characteristics of lymphocytic myocarditis (Figure 3A). Haematoxylin and eosin staining showed fibrosis and interstitial oedema. Immunostaining demonstrated numerous CD3-positive lymphocytes. CMR (MAGNETOM Vida 3.0T; Siemens, Munich, Germany) on day 34 showed a high native T1 value (Figure 1 and Video S3).

The patient was a 48 year-old woman without any past medical history. She was not vaccinated against COVID-19. Six days before admission, she developed a fever and tested positive for COVID-19 by antigen test (Table 1). She arrived at our hospital for fatigue and dyspnoea. She was in a state of cardiogenic shock. A 12-lead ECG showed STE in leads II, III, aV_F and V_4–6 (Figure 1). TTE demonstrated diffuse LV dysfunction with an LVEF of 30% and pericardial effusion. There was oedematous thickening of the RV wall, which was particularly evident at the tricuspid valve annulus (Figure 4, Videos S4, S5, and S6). Cardiac enzymes were elevated (Table 2). A slow-flow phenomenon (collected Thrombolysis in Myocardial Infarction frame counts⁶: LAD 117, LCX 54, RCA 32) without significant stenosis was observed by CAG. EMB from RV was performed. Inotropes were initiated, but the patient's condition continued to deteriorate. TTE confirmed progressive LV dysfunction (LVEF 10%) and RV dysfunction. On day 2, the patient required MCS [VA-ECMO and a percutaneous LV assist device (ImpellaCP, Abiomed, MA, USA)]. Heparin was administered as the ImpellaCP purge solution. She was treated with methylprednisolone (1000 mg for 3 days) and remdesivir (200 mg for 5 days). Cardiac function started to improve on day 5. Pericardiocentesis was performed for cardiac tamponade on day 7. Approximately 100 mL of pale-yellow pericardial fluid was aspirated, and haemodynamics improved immediately after the procedure. On day 8, she was weaned off MCS. She was discharged on day 33. Although cardiac function had improved to the normal range, she continued to experience post-exertional malaise and dyspnoea. A cardiopulmonary exercise test (CPX) conducted on day 72 revealed reduced exercise capacity with percent peak oxygen uptake (%peakVO₂) at 74.2%, prompting the initiation of the cardiac rehabilitation program. However, despite a 3 month rehabilitation program, a CPX on day 164 revealed a further decline of % peak VO₂ (68.7%).

The EMB revealed MVT and interstitial oedema without evidence of myocarditis (Figure 3B). Myocardial cells exhibited cytoplasmic vacuolation, suggesting microcirculatory ischaemia. The second EMB demonstrated improvement in pathological findings (Figure S1). CMR on day 19 showed no typical late gadolinium enhancement but revealed high native T1/T2 values (Figure 1 and Video S7) and an extracellular volume (ECV) of 37%. CMR on day 96 showed decreased native T1/T2 values and ECV (32%) (Figure S2 and Video S8).

Here, we describe two cases of COVID-19 patients who developed acute fatal cardiac dysfunction and myocardial injury, leading to cardiogenic shock. The first case had typical pathological features of lymphocytic myocarditis. The second case lacked evidence of myocarditis and showed MVT associated with myocardial ischaemia and marked interstitial oedema.

The aetiology of patients with suspected COVID-19-myocarditis is not established. The incidence of pathologically proven myocarditis ranges from 2%–30% in autopsied patients with COVID-19 myocardial injury.^2-4 Some cases showed pathological patterns of diffuse interstitial macrophage infiltrate and micro-thrombi,^{3, 4} notably distinct from the dense lymphocytic infiltrate of typical viral myocarditis. Other investigations also reported the incidence of microthrombi in autopsied hearts with COVID-19 to be as high as 64%–70%.^{5, 7} Additionally, among cases of patients with suspected FM after COVID-19 evaluated pathologically, relatively few showed typical features of lymphocytic myocarditis. Other cases showed mild inflammatory infiltration predominantly with macrophage or MVT (Tables S1 and S2). These findings suggested that the common pathogenesis of suspected COVID-19-related FM is not lymphocytic myocarditis. Dense lymphocytic infiltration directly degenerates and damages myocytes in typical viral myocarditis. A different mechanism is considered in myocardial injury by MVT. MVT can contribute to global myocardial ischaemia, causing diffuse cardiac dysfunction, as pathologically shown by coagulation necrosis and diffuse cytoplasmic vacuolization of myocytes, both characteristic of acute ischaemia. Patients with COVID-19 are potentially at high risk for thromboembolic events resulting from coagulopathy,⁸ cytokine storm, and direct actions on platelets and the endothelium. SARS-CoV-2 promotes endothelial complement deposition and mainly C5a production,⁹ which activates neutrophils, leading to tissue factor (TF) release and the formation of neutrophil extracellular traps (NETs).⁸ Thrombin generated by TF activates platelets while NETs containing C3a, histones and other platelet-activating substances further enhance platelet activation. This creates a positive feedback loop where activated platelets promote more NET formation. Furthermore, cytokines like IL-8, TNFα and IL-6 stimulate endothelial release of unusually large von Willebrand factor multimers,¹⁰ which accumulate and promote micro-thrombi formation due to reduced ADAMTS13 activity. In addition, it is suggested that SARS-CoV-2 directly invades cardiomyocytes, leading to cardiac dysfunction by impairing ACE2 activity, reducing angiotensin-(1-7)/MAS cardioprotective effects and increasing angiotensin II levels, which worsens cardiotoxicity.¹¹

The clinical course is similar in both cases, but there were some differences. Previously reported cases of COVID-19-lymphocytic FM have progressed to cardiogenic shock more than a week after COVID-19 infection (Table S1). Patient 1 may have had an asymptomatic infection when a family member had a COVID-19 infection a few weeks earlier. Cases without pathological lymphocytic infiltration, including our MVT case, deteriorated rapidly within a week (Table S2). Also, in the MVT case, a slow-flow phenomenon was noted in CAG, suggesting increased microvascular resistance. Echocardiography in the MVT case showed pronounced abnormalities in the right heart and tricuspid valve annulus, and these findings resolved as cardiac function recovered. Previous CMR studies revealed high native T1 and T2 signal intensity and RV dysfunction in COVID-19 patients with cardiac symptoms,¹² which was more pronounced in the MVT case, presumably due to the extensive cardiac injury.

Various treatment was selected in previously reported suspected COVID-19-related FM (Tables S1 and S2). Given that recovery of cardiac function has been reported in most cases that survived to discharge, management of cardiogenic shock is crucial. Anticoagulants, particularly heparin, are considered essential and have been reported to be effective against COVID-19-related thrombosis.¹³ Anticoagulants may have affected the dissolution of MVT, as observed by a serial EMB. The steroids have been reported to suppress the cytokine storm in COVID-19 and reduce mortality¹⁴ and may also be effective in inhibiting MVT. However, despite medical treatments and improved cardiac function, the MVT case continued to experience fatigue and decreased exercise capacity. This did not improve even after 3 months of cardiac rehabilitation program, suggesting the involvement of post-acute sequelae of COVID-19. Long-COVID has been reported to be associated with persistent endothelial dysfunction and MVT following acute COVID-19.¹⁵ It is suggested that endothelial dysfunction and ischaemic damage due to MVT extend beyond the cardiac function to affect other systemic organs. While effective treatments for Long-COVID are still not established, earlier anticoagulation therapy may improve symptoms.¹⁵ Also, long-term monitoring may be essential for patients after recovery from severe COVID-19 myocardial injury. Rehabilitation after recovery from a critical condition is crucial; however, exercise has been suggested to be potentially harmful or less effective in Long-COVID cases.¹⁵ Therefore, it is essential to conduct rehabilitation under careful supervision. While this report provides valuable insights into the potential cardiac manifestations of COVID-19, it is important to acknowledge the inherent limitations of a case report, including the lack of a control group, potential selection bias and possible diagnostic challenges despite advanced imaging and biopsy techniques.

In cases of myocardial injury and cardiogenic shock associated with COVID-19, it is essential to consider the involvement of both FM and MVT. The clinical course in the acute phase is similar, but patients with MVT may have long-lasting effects beyond cardiac function that affect systemic organs, even after the acute phase of the disease has resolved. Further investigation and a more extensive series of patients with COVID-19 myocardial injury are needed to determine the treatment and how frequently MVT contributes to myocardial damage in the absence of coronary epicardial thrombosis.

None declared.