Diffuse mononuclear inflammatory response to COVID-19: friendly fire or smoldering enemy?

A recent paper by Goldman et al. [1] described a series of six patients who presented to the University of Rochester Medical Center over a 2 years period with acute onset of cardiogenic shock. All six patients underwent endomyocardial biopsy for pathologic diagnosis and every biopsy demonstrated interstitial myocardial inflammation that consisted of predominantly large, mononuclear cells. Five of the six cases (5/6) showed microvascular luminal narrowing associated with reactive mononuclear cells, changes similar to antibody mediated rejection in cardiac allografts. The biopsies were subsequently stained for cluster of differentiation (CD[68]) and CD31 which showed CD68 positive cells adjacent to and sometimes within myocardial microvasculature. The authors proposed a new term, distinct from currently used classification systems, to characterize the findings.

According to the World Health Organization and/or International Society and Federation of Cardiology Task Force on the Definition and Classification of Cardiomyopathies, inflammatory cardiomyopathy is defined as myocarditis in association with cardiac dysfunction. The term myocarditis is defined as “an inflammatory disease of the myocardium” that is diagnosed by established histological, immunological, and immunohistochemical criteria [2]. In the position statement of the European Society of Cardiology, they cite the World Health Organization as stating that an endomyocardial biopsy could be diagnosed as myocarditis after immunohistochemical detection of focal or diffuse mononuclear infiltrates with >14 leukocytes per 1 mm2 (CD3+ T lymphocytes and/or CD68+ macrophages). The consortium further refined the definition to include ≥14 leucocytes/mm2, including up to 4 monocytes/mm2, with the presence of CD3-positive T-lymphocytes at ≥7 cells/mm2 [3]. The Dallas Criteria, generally considered the gold standard for endomyocardial diagnosis of myocarditis, simply state that the diagnosis requires the presence of an inflammatory infiltrate and definitive myocyte damage [4]. Some reports have shown large numbers of macrophages infiltrating the myocardium but these cases are usually discussed in the context of chronic myocarditis which is associated with significant fibrosis. By these currently accepted criteria, the Goldman cases would not qualify as acute myocarditis because they require the presence of both CD3+ T lymphocytes independent from the presence of macrophages and the macrophages present do not cluster at >14 leukocytes per 1 mm2. In their analysis, Goldman noted that two cases occurred in patients with no significant prior medical history while two others developed in the background of known systemic lupus erythematosus (SLE); one of which was associated with macrophage activation syndrome . They concluded that all six cases represented a form of histiocytic inflammatory myocardial disease distinct from chronic inflammatory cardiomyopathies and coined the term “histiocytic myocardial inflammatory disease” to explain the findings of acute inflammation with associated cardiac dysfunction. They further hypothesized that histiocytic myocardial inflammatory disease may be primary or secondary and that myocyte necrosis could be seen in both forms.

We recently published a paper in this Journal also describing the diffuse interstitial monocyte and/or macrophage myocardial infiltrate present in autopsied patients who died of severe COVID-19 [5]. In our study, we noted that SARS-CoV-2 infected hearts did not show a significant lymphocytic infiltrate characteristic of acute viral myocarditis which was commonly assumed to be underlying cardiac complications in early reports from China. We, as well as others, [6] noticed that some small capillaries contained platelet/fibrin clots associated with varying degrees of acute myocyte necrosis in nearby myocytes. We originally hypothesized [7] that SARS-CoV-2 was infecting the endothelium causing damage and the reported swelling and clot formation. After recognizing the increased monocyte and/or macrophage population, we proposed that this pattern of inflammation may be a unique consequence of COVID-19 infection. Given the cases reported by Golden et al., we agree that acute macrophage-predominant inflammatory infiltrates may represent a distinct form of inflammatory heart disease that occurs in a variety of clinical contexts and may be important in understanding the growing concern regarding long COVID-related cardiac effects.

Increased numbers of macrophages in COVID hearts were reported early by Basso et al. [8]. The European consortium found that 86% of the 21 COVID hearts they examined contained increased numbers of interstitial macrophages. They concluded that the increased macrophages likely reflected the underlying vascular damage consequent to the co-morbid vascular conditions of the patient population. Since hypertension, diabetes, and diabetes are known to cause endothelial dysfunction, it was important to control for these underlying conditions. Our study showed that COVID-19 hearts contained increased numbers of CD68+ cells compared to control, non-COVID autopsied hearts from patients with these important underlying risk factors [5].

Goldman also described coagulative myocyte necrosis in two of the six cases based on Masson trichrome staining. Interestingly, the same two hearts showed cytoplasmic positivity with C4d and/or C9 and the authors concluded that these areas represented coagulation necrosis with complement activation. It was not clear to the authors whether the coagulative necrosis was secondary to direct histiocytic injury (complement mediated cell death) or to ischemia and/or reperfusion secondary to microvascular injury and resulting hypoperfusion. Pellegrini et al. [6] performed a systematic pathological analysis of 40 hearts from hospitalized patients dying of COVID-19 in Italy. Of the 40 hearts examined, 35% (14) contained myocyte necrosis, predominantly in the left ventricle. Importantly, the incidence of severe coronary artery disease (defined as >75% cross-sectional narrowing) was not different between hearts with or without myocyte necrosis. Cardiac thrombi were present in 11 of the 14 hearts with necrosis (78.6%) and the microthrombi contained significantly more fibrin and terminal complement (C5b-9) immunostaining than thromboemboli from COVID-19 negative patients. The authors concluded that microthrombi was the most common cause of myocyte necrosis in COVID patients.

In our cohort, we observed focal myocyte positivity with C5b-9 in individual myocyte necrosis consistent with the data from Goldman. However, in the larger areas of coagulative necrosis, the dead myocytes were negative for C4d and C5b-9 staining. Similar to other reports, the majority of necrotic myocytes were present in the left ventricle but were also identified in the right ventricle which may be consistent with acute right heart strain commonly seen in severely ill COVID-19 patients. However, because many of the necrotic myocytes showed evidence of contraction band necrosis, we believe that the majority injury is due to microvascular and/or endothelial injury and consequent ischemia and/or reperfusion injury. Interestingly, a recent report by Dutt et al. demonstrates that enhanced circulating endothelial cells (CD452CD31+ CD34+ CD146+), circulating endothelial progenitors (CD452CD31+ CD34+/2CD1462), and neutrophils (CD11b+ CD66b+) correlated with the severity of COVID-19 disease suggesting that SARS-CoV-2 infection causes damage and increased turnover of endothelial cells [9]. In our recent review article, we pointed out a possible important role of monocyte-derived macrophages in the pathogenesis of severe COVID-19 induced myocardial damage either by causing a cytokine-mediated endothelial damage and/or by causing localized endothelial damage and/or apoptosis [10].

Diffuse myocardial infiltration with mononuclear inflammatory cells has been typically discussed in relation to chronic myocarditis and/or inflammatory dilated cardiomyopathy [11]. Interestingly, diffuse interstitial mononuclear infiltrates have been shown in other systemic viral infections (e.g., Hentavirus, Ebola) and was even described in the SARS-Co-V epidemic of 2009 [12]. There is now convincing evidence that monocyte/macrophages can be infected by SARS-CoV-2 which suggests that the virus can reach the heart during transient viremia and/or through actual infiltration of the myocardium [13, 14]. It is important to note that changes described in most COVID-19 autopsy hearts may not exactly coincide with currently accepted definition of inflammatory cardiomyopathy. Although there have been reports of cardiac and renal dysfunction associated with SARS-CoV-2 infection (excluding rare reports of acute viral myocarditis) [15, 16], respiratory damage and/or failure is responsible for the vast majority of COVID-19 mortality. Furthermore, to date no documented reports of long-term cardiac dysfunction and/or ventricular remodeling in patients who survive COVID-19.

Given the multiple reports of diffuse monocyte and/or macrophage infiltration of COVID-19 hearts, it is interesting to speculate whether this pattern of inflammation is “friendly fire,” given that the non-traditional monocyte and/or macrophage subsets are considered to be anti-inflammatory in nature and therefore protective of the heart, or “enemy fire” in that persistent presence of significant numbers of inflammatory cells could result in chronic, low grade inflammation and potentially in fibrosis and/or remodeling of the affected ventricle. At this point, there is not enough basic knowledge (i.e., immunology/microbiology) or clinical follow up to answer these questions. It is clear that more investigation is needed to understand the effects of COVID-19 on the heart especially given the increasing concern about cardiac symptoms in COVID survivors (“long haulers”). These studies should include both basic and clinical studies but should include the systematic collection of endomyocardial biopsies and autopsied hearts to generate histology and tissue biobanks to better characterize the pathology of SARS-CoV-2 infection in the heart.