Recovery of systemic hyperinflammation in patients with severe SARS-CoV-2 infection

Abstract Introduction Patients with cardiovascular disease (CVD) and acute SARS-CoV-2 infection might show an altered immune response during COVID-19. Material and methods Twenty-three patients with CVD and SARS-CoV-2 infection were prospectively enrolled and received a cardiological assessment at study entry and during follow-up visit. Inclusion criteria of our study were age older than 18 years, presence of CVD, and acute SARS-CoV-2 infection. The median age of the patient cohort was 69 (IQR 55–79) years. 12 (52.2%) patients were men. Peripheral monocytes and chemokine/cytokine profiles were analysed. Results Numbers of classical and non-classical monocytes were significantly decreased during acute SARS-CoV-2 infection compared to 3-month recovery. While classical monocytes reached the expected level in peripheral blood after 3 months, the number of non-classical monocytes remained significantly reduced. Discussion All three monocyte subsets exhibited changes of established adhesion and activation markers. Interestingly, they also expressed higher levels of pro-inflammatory cytokines like macrophage migration inhibitory factor (MIF) at the time of recovery, although MIF was only slightly increased during the acute phase. Conclusion Changes of monocyte phenotypes and increased MIF expression after 3-month recovery from acute SARS-CoV-2 infection may indicate persistent, possibly long-lasting, pro-inflammatory monocyte function in CVD patients. GRAPHICAL ABSTRACT


Introduction
In the past two years, the COVID-19 pandemic has significantly threatened the health system worldwide. It has been rapidly recognised that patients with cardiovascular disease (CVD) or associated risk factors such as arterial hypertension, diabetes mellitus, renal failure, and obesity are at high risk for an unfavourable course of COVID-19 (Huang et al. 2020, Zhou et al. 2020. COVID-19 has been associated with severe systemic inflammatory response and cytokine storm leading to thrombotic complications, microcirculatory disturbances and organ failure (Huang et al. 2020, Zhou et al. 2020, Jose et al. 2020. Thus, SARS-CoV-2 infection may have a long-term impact not only on the cardiovascular system but also on the immune response (Xiong et al. 2020, Zheng et al. 2020 Force for the management of C-otESoC 2022). Recently, we found that patients with CVD are at increased risk of life-threatening heart and lung injury due to pro-inflammatory and prothrombotic responses during SARS-CoV-2 infection (Mueller et al. 2021). We identified phenotype changes of circulating monocyte subtypes together with reduced numbers of circulating non-classical CD14 dim CD16 + monocytes and sequestration in affected organs precedes and predicts rapidly progressive respiratory and multiorgan failure (Mueller et al. 2021) Further, we and others found that platelet activation and hyperinflammation is associated with prognosis of patients with CVD and COVID-19 (Gu et al. 2021, Langnau et al. 2021. SARS-CoV-2 infection not only activates the antiviral immune response but, in addition, is responsible for uncontrolled immune responses in a number of cases, particularly increased release of pro-inflammatory cytokines. This in turn might induce lymphopenia, lymphocyte, granulocyte, and monocyte dysfunction and reduce immune cell numbers of clinical relevance (Yang et al. 2020). Another study showed that the numbers of intermediate monocytes producing interleukin (IL-) 6 were increased in patients that were admitted to intensive care unit (ICU) compared to patients with less severe courses of COVID-19 (Merad et al. 2020). Additionally, a significantly reduced amount of peripheral non-classical monocytes predicted an adverse clinical outcome in patients with CVD and severe SARS-CoV-2 infection (Mueller et al. 2021). To date, the long-term effects of SARS-CoV-2 infection have not been fully understood. Therefore, it is important to investigate the immune responses and immunophenotypes during acute phases and after recovery from SARS-CoV-2 infection. Along with clinically relevant complications and consequences such as fatigue, muscle weakness, dyspnoea, arrhythmias, anxiety, and depression, it is essential to investigate the long-term outcome of severe SARS-CoV-2 infection on immune cells and immune responses (Huang et al. 2021). The aim of the present study was to characterise the long-term changes of systemic inflammation in patients with severe acute SARS-CoV-2 infection on immune cells. We compared subgroups of circulating monocytes in the same group of COVID-19 patients in the acute phase of hyperinflammation and after 3-months recovery follow-up as well as a healthy control group using immunophenotyping by multicolour flow cytometry.

Study design, study populations and inclusion criteria
In this study, we consecutively enrolled 23 patients at the Department of Cardiology and Angiology of the University Hospital Tübingen, Germany from February 2020 until September 2020. Twenty-three patients with pre-existing CAD and an acute, severe SARS-CoV-2 infection were included within 12 hours after hospital admission. As control, 40 healthy controls without SARS-CoV-2 infection were included in our study. The healthy control group showed a median age of 41 (IQR 28-53), 41% were men. Healthy controls did not suffer from any disease and did not take any relevant medication, especially no medication on daily basis. All 23 patients were analysed after a recovery period of 3 months in our department. None were lost to follow-up. All patients received a clinical and cardiac examination including echocardiography, electrocardiography, concomitant medication, comorbidities and blood sampling for routine laboratory parameters, marker expression on monocytes, and chemokine profiling. From nasopharyngeal secretions SARS-CoV-2 infection was confirmed by real-time reverse transcriptase polymerase chain reaction (PCR). Inclusion criteria of our study were age older than 18 years, confirmed CVD or CV risk factors, and acute SARS-CoV-2 infection. Exclusion criteria were other microbial infections. The study was approved by the local ethics committee (240/2018B02) and complies with the declaration of Helsinki and the good clinical practice guidelines on the approximation of the laws, regulations and administrative provisions of the member states relating to the implementation of good clinical practice in the conduct of clinical trials on medicinal products for human use. Written informed consent was obtained from every patient.
We determined N-terminal-pro-B-type natriuretic peptide (NT-pro-BNP, >300 ng/L), high sensitive troponin I (hs TNI, >37 ng/L), IL-6, and C-reactive protein (CRP, >0.5 mg/dL) as elevated laboratory markers of myocardial and inflammatory distress. As echocardiographic parameters the left and right ventricular function, right ventricular dilatation, presence of tricuspid valve regurgitation, and pericardial effusion were included according to current guidelines (Lang et al. 2015, Parasuraman et al. 2016).

Measurement of plasma levels of cytokines and chemokines (LEGENDPlex)
The  Table S3) (BioLegend, San Diego, California, USA). The assays were performed according to the manufacturer's manual. FACS Lyric (BD Biosciences, Franklin Lakes, New Jersey, USA) was used for the measurement and data analysis was performed with the LEGENDPlex Data Analysis Software (BioLegend, San Diego, California, USA).

Statistical analysis
Statistical analysis of participants' clinical and laboratory baseline characteristics in relation to measured monocyte and platelet phenotypes and marker expression was performed. Non-normally distributed continuous data are represented as median with interquartile range (IQR) and normally distributed continuous data are represented as mean with standard deviation (SD). Two group comparisons for non-normally distributed continuous variables were performed using a Mann-Whitney U test while normally distributed continuous variables were compared using Student's T-test. Paired non-normally distributed continuous variables were compared using Wilcoxon test. Three group comparisons for non-normally distributed continuous variables were performed using a Kruskal Wallis test while normally distributed continuous variables were compared using One Way ANOVA. Categorical variables are represented as total numbers and proportions of participants and comparison was performed using chi-square test. Comparisons were considered statistically significant if two-sided p value was <0.05. All statistical analysis was performed with IBM SPSS Statistics software version 26 (SPSS, Inc.) and GraphPad Prism Version 8.4.0 (GraphPad Software). Unsupervised data analysis was done using OMIQ data analysis software (Omiq inc., Santa Clara, CA, USA). Heatmaps were generated with the metaclusters obtained from PhenoGraph and clustered hierarchically on all surface markers with a euclidean distance metric to indicate the similarity of the populations.

Clinical characteristics of patients with severe SARS-CoV-2 infection
From February to September 2020, we prospectively studied a consecutive cohort of 23 patients at cardiovascular risk or with cardiovascular disease (CVD), who were treated for severe respiratory failure associated with acute SARS-CoV-2 infection in our hospital (Supplemental Table S1). 40 healthy participants served as controls. The healthy control group showed a median age of 41 (IQR 28-53), 41% were men. Healthy controls did not suffer from any disease and did not take any relevant medication, especially no medication on daily basis. Baseline characteristics and demographics of the overall cohort are given in Supplemental Table S1. The median age of the population was 69 (IQR 55-79) years. 12 (52.2%) patients were men. 11 (47.8%) patients presented with a Horovitz Index (HI) < 300 mmHg on admission (Supplemental Table S1). In the acute COVID-19 phase most prominent clinical signs were cough, dyspnoea, and fevers. All patients revealed radiological signs associated with their respiratory failure in the X-ray of the chest (Table 1). Pericardial and/or pleural effusion were present in 11 (48%) and 7 (30%) patients, respectively (Table 1). While there was no clinically relevant leukocytosis documented upon admission, we found lymphocytopenia of 660/µL (530-1150), elevated CRP and IL-6 levels of 2.5 mg/dL (1.3-12.5) and 27.3 (9.5-37.4), respectively (Table 1). Furthermore, NT-pro-BNP and D-Dimer were increased with 323 ng/L (132.5-825.5) and 0.9 µg/dL (0.6-1.9), respectively (Table 1). During their hospital stay, 9 (39.1%) patients were admitted to the intensive care unit (ICU) due to progressive respiratory failure requiring transient high-flow oxygen therapy (n = 4) or mechanical ventilation (n = 5) (Supplemental Table S1). At 3-months follow-up 43.5% of patients reported persistent dyspnoea as most prominent clinical symptom. Furthermore, one patient was re-hospitalized for acute Non-ST-elevation myocardial infarction during follow-up. 5 patients complained of chronic fatigue. No events of thrombosis, stroke, bleeding, or rhythm disturbances were evident (Supplemental Tables S1 and S2). In most patients pericardial and pleural effusion were absent in the recovery phase and radiological pulmonary pathologies recuperated (Table 1). Furthermore, initially impaired laboratory parameters including lymphocyte count, elevated CRP-, IL6-, NT-pro-BNP-, creatine kinase-(CK), lactate dehydrogenase-(LDH) levels during acute infection reached normal values at the follow up visit after three months ( Table 1).
In line with our previous findings (Mueller et al. 2021), three months after SARS-CoV-2 infection the numbers of circulating CD14 dim CD16 + non-classical monocytes was dramatically reduced in patients with acute SARS-CoV-2 infection and returned to near-normal levels three months later (CD14 dim CD16 + non-classical monocytes; acute vs post vs healthy control; median + IQR; 1640 (526.3-6110) vs 12190 (2420-30020) vs 21905 (13883-36360); acute vs post p = 0.0013; acute vs healthy p < 0.0001; post vs healthy p = 0.0701, Figure 1D). Similarly, the number of circulating CD14 + CD16classical monocytes that were also slightly reduced during the acute phase of the infection and returned to normal levels after 3-months recovery (CD14 + CD16classical monocytes; acute vs post vs healthy control; median + IQR; 193070 (122800-321180) vs 340280 (285140-491200) vs 292455 (240380-393835); acute vs post p = 0.0081; acute vs healthy p = 0.0779; post vs healthy p = 0.746, Figure 1D) whereas the numbers of intermediate monocytes (CD14 + CD16 + ) remained unchanged throughout the infection ( Figure 1D). Previously, we found that the function of circulating monocytes was impaired in CVD patients with severe acute SARS-CoV-2 infection compared to CVD patients without SARS-CoV-2 infection, as indicated by low expression of CD62L, CX3CR1 and HLA-DR as established markers of adhesion, migration, and T-cell activation (Mueller et al. 2021). Here, we additionally assessed the expression of markers on subtypes of monocytes     vs post p > 0.9999; acute vs healthy p = 0.0013; post vs healthy p = 0.0004 ( Figure 2D).
For deeper characterisation of the immunophenotypes, we also analysed the intracellular expression of the chemokines CXCL12 and macrophage migration inhibitory factor (MIF) in all three subtypes of circulating monocyte. CXCL12 and MIF play important roles in inflammatory and atherosclerotic processes via chemotactic regulation of lymphocytes and monocytes (Calandra et al. 2003, van der Vorst et al. 2015, Gao et al. 2019. While CXCL12 expression was significantly increased in all monocyte subsets from patients in the acute phase of the infection and reached levels comparable to healthy controls after 3-months recovery ( Figure 2F).
These results indicate that monocyte chemotaxis, adhesion, migration, and B/T cell activation capabilities are influenced by acute SARS-CoV-2 infection and may recover three months after acute infection. However, several alterations of monocyte function do not recover during this three-month period back to expression levels of healthy controls (Figure 2). For a direct comparison of each patient during acute phase of infection and a recovery of 3 months, we prepared new figures for each cell count and investigated marker (Supplemental Figures S2 and S3).

Hierarchical clustering of monocyte subtypes identifies specific phenotypes in acute and recovery phase of SARS-CoV-2 infection
In addition to the manual gating strategy of multicolour-flow cytometry described above, we performed unsupervised data analysis by first applying uniform manifold approximation and projection (UMAP) (McInnes et al. 2018) dimension reduction of gated monocytes to group phenotypically similar events (Figures 3 and 4). This was followed by unsupervised clustering analysis using Phenograph (Levine et al. 2015), which uncovered distinct changes in monocyte marker profiles (Figures 3 and 4). Activation and migration surface markers such as CCR2 and CD11a are increased during acute phase of infection in cluster MT23, MT17 and MT14 comparable to our manual gating strategy (Figure 3). Unsupervised clustering with PhenoGraph resolved 28 clusters (MT01-MT28) for the monocyte chemokine receptor flow cytometry panel, of which 28 showed significant differences regarding the frequency of the clusters MT01-MT28 between patients with acute SARS-CoV-2 infection and patients after 3-month recovery. Only relevant significant clusters are shown in Figure 4. Here only seven relevant clusters are shown. Cluster MT03 and MT24 showing increased expression of intracellular CXCL12 and less CXCL14 in the acute phase compared to the recovery phase of SARS-CoV-2 infection, while MT27 showed an increase of both: CXCL12 and CXCL14 in patients during acute infection. Similar to our manual gating, the expression of CD11b is reduced in monocytes after 3-month of recovery in cluster MT27, MT14, MT03 and MT24 compared to the acute phase of infection. Moreover, we found that the median expression of monocytes shown by CD14 and CD16 surface expression was increased during the acute phase compared to the recovery phase in several clusters (Figure 4). Unsupervised OMIQ analysis confirmed our data obtained by our manual gating strategy.

Plasma levels of MIF and other pro-inflammatory cytokines in patients with acute SARS-CoV-2 infection were significantly increased in CAD patients
Next, we asked whether the assessment of systemic inflammation allows to characterise the acute phase of hyperinflammation compared to the recovery phase. We analysed plasma of CVD patient with acute SARS-CoV-2 infection and 3-months recovery with an inflammation panel of 26 chemokines/cytokines. The cytokines and chemokines were analysed as described in the method section. In particular, we found differences of the pro-inflammatory mediators CXCL10, IL-18, CCL2, IL-6, IL-8, IFN-γ, and CXCL11 which were significantly elevated in the plasma of CVD patient with acute SARS-CoV-2 infection compared to patients with 3-months recovery. Interestingly, also the cytokine MIF was significantly increased during the acute phase of infection ( Figure 5A). MIF exhibit chemokine-like functions and is related to several acute and chronic inflammatory diseases as well as cardiovascular disease (Morand et al. 2006, Zernecke et al. 2008, Muller et al. 2014, Muller et al. 2015. Further, MIF is able to function as a pro-atherogenic factor and promote monocytes adhesion, migration and differentiation by interacting with CXCR2 and CXCR4 (Bernhagen et al. 2007, Cheng et al. 2010, Muller et al. 2015 Interestingly, IL-33, CXCL5, CCL11, CXCL1, IL-17A, CCL20, and CCL4 revealed significantly higher levels at 3-months of recovery compared to the acute phase ( Figure 5B). Other tested mediators comprising IL-10, IFN-α, IL-23, IL-1ß, CXCL8, CCL5, CCL17, TNF-α, CXCL9, IL-12p70, and CCL3 showed similar plasma levels in both groups ( Figure 5A and B).  We observed a significant increase of pro-inflammatory chemokines/cytokines like MIF, CCL2, IL-6 and IFN-y during the acute phase of infection compared to the timepoint of 3-months recovery.
Our data indicate that significant changes of circulating monocytes and chemokines occur in the acute and recovery phase of SARS-CoV-2 infection in CVD patients. The increased numbers of intracellular MIF expression in classical, intermediate, and non-classical monocytes and elevation of distinct chemokines in the recovery phase imply a sustained immune response following SARS-CoV-2 infection. Thus, it is tempting to speculate that monocytes are prominent inflammatory cells involved in the acute defense of the infectious disease and prominent changes of monocyte subgroups in the recovery phase may contribute to the prolonged immune response following SARS-CoV-2 infection.
COVID-19 has threatened our societies worldwide during the last 2 years. There is an unbelievable amount of scientific data available that disclosed the underlying pathophysiology in patients with SARS-CoV-2 infection. One major prominent aspect of severe SARS-CoV-2 infection is the systemic hyperinflammation associated with respiratory stress syndrome, microcirculatory dysfunction, and organ failure. Recently, we found that activation of monocytes occurs in severe COVID-19 with an enhanced pulmonary sequestration of especially non-classical (CD14 dim CD16 + ) monocytes (Mueller et al. 2021, Chilunda et al. 2021, Degauque et al. 2021. These data suggested that monocyte contribute significantly to respiratory distress and organ failure in COVID-19.
The purpose of the present study was to characterise changes of monocyte subtypes and relevant plasma chemokines during the recovery phase of COVID-19 at 3-months follow-up in patients with cardiovascular disease. We focussed on CVD patients in our study because CVD, CAD patients or patients with cardiovascular comorbidities have an increased risk for life-threatening heart and lung injury and organ failure (Oren et al., Task Force for the management of C-otESoC 2022). SARS-CoV-2 infection is characterised by pro-inflammatory and pro-thrombotic events causing in CVD patients acute coronary syndrome with subsequent impairment of left or right ventricular function (Xiong et al. 2020, Tang et al.). Further, acute right heart failure occurs in CVD or CAD patients due to pre-existing right heart and diastolic dysfunction (Oren et al., Task Force for the management of C-otESoC 2022). Pneumatic infiltrates and lung involvement cause right heart dysfunction upon SARS-CoV-2 infection (Xiong et al. 2020). Furthermore, infiltration of inflammatory immune cells like monocytes or macrophages in CAD patients have been related to atherosclerotic lesions due to increased expression of cytokines (Woollard et al. 2010, Swirski et al. 2018, Wolf et al. 2019.
At 3-months follow-up almost half of the analysed patients (43.5%) reported persistent dyspnoea as most prominent clinical symptom and as a sign of possible long-COVID. Five patients complained of chronic fatigue. 56.5% of the enrolled patients revealed a complete clinical recovery from initially severe COVID-19 at the follow-up visit and were free from cardiovascular and neurological symptoms. Furthermore, we found that changes of monocyte subgroups were highly dynamic when comparing the acute to the recovery phase. One prominent finding was that the expression of MIF in monocytes was not altered in the acute phase of SARS-CoV-2 infection but increased 3-months later. MIF is an ubiquitously expressed chemokine-like protein which was found to be an important regulator of innate immunity (Calandra et al. 2003). Binding of MIF to cells promotes NF-κB activation and inflammation, migration or survival (Calandra et al. 2003, Kim et al. 2017. Our data further emphasise the role of MIF and circulating monocytes in the immuno-defense following viral infections such as SARS-CoV-2. A sustained expression of MIF in circulating monocytes may contribute to the cellular-based immuno-defense system in COVID-19. Monocytes are innate blood cells that are early cellular responders in acute infection (Jakubzick et al. 2017). Whereas classical monocytes (CD14 ++ CD16 -) are critical for the initial inflammatory response, nonclassical (CD14 dim CD16 + ) "patrolling" monocytes are commonly viewed as anti-inflammatory guardians of the innate immune system (Buscher et al. 2017). Our results are in line with the current literature showing that circulating monocyte subsets return to normal proportions after a follow-up of six months upon infection (Utrero-Rico et al. 2021). Most of the literature are focussing on immune phenotyping of monocyte and macrophages during the acute phase of infection. Studies show that inflammatory monocytes are hyperactivated and secreting large amounts of pro-inflammatory cytokines (Merad et al. 2020, Knoll et al. 2021. Thus, it is tempting to speculate, that especially MIF-positive monocytes are an imprint of severe SARS-CoV-2 infection that may contribute to immunity and re-infection. This conclusion is further substantiated by the fact that a variety of chemokines (e.g., IL-33, CCL17, CXCL1, IL-17A) were elevated months after SARS-CoV-2 infection. IL-33 is a cytokine that induces of T-helper 2 (Th2) cell priming (Komai-Koma et al. 2007). CCL17 is chemotactic for T-regulatory cells. CXCL1 serves as chemoattractant for neutrophils (Sawant et al. 2016, Metzemaekers et al. 2020. IL-17A has been implicated in immune response to infectious pathogens (Ge et al. 2020). Thereby, our observations indicate that these elevated cytokine/chemokine levels were also associated with altered and even partly enhanced immune response even after a 3-months recovery phase after COVID-19. However, at present we cannot provide evidence that here observed alterations of plasma chemokines and subsets of circulating monocytes play a role in a sustained protection for re-infection with SARS-CoV-2. We conclude that patients with CVD and acute SARS-CoV-2 infection showed changes regarding their phenotype of monocytes and their chemokine profile after 3-months recovery. Altered monocyte function and increased MIF expression was characteristic for the recovery phase. One can speculate that MIF expression may serve as additional biomarker to identify patients at risk for an altered, possibly prolonged immune response after acute SARS-CoV-2 infection.
We are aware that our study is rather observational and hypothesis-generating, especially as the analysis is limited by the small patient number in the observed group. Furthermore, we can only provide data after 3-months of follow-up and cannot provide data after 6 or 12 months. Therefore, our data are also limited regarding their impact on long-COVID. Therefore, large scale randomised studies are essential for validation of our observations on the predictive value of MIF and other pro-inflammatory cytokines regarding the recovery of the immune response after acute COVID-19infection as well as their impact to predict patients at risk for an unfavourable course of COVID-19 and long-COVID.

Conclusion
We found that the changes in immunophenotypes of circulating monocytes and plasma levels of most chemokines/ cytokines that occurred or increased during the acute phase of severe SARS-CoV-2 infection normalised after a 3-month recovery. However, we also found that plasma levels of certain chemokines, in particular MIF, were still or newly elevated 3-months post COVID-19 compared to the acute infection phase, indicating a persistent alteration of the immune response after COVID-19 recovery.

Disclosure statement
No potential conflict of interest was reported by the authors.

Funding
This work was supported by the German Research Foundation (DFG) -Project number 374031971-TRR 240 and by the Ministry of Science, Research and the Arts of the State of Baden-Württemberg (COVID-19 Funding) grant to Meinrad Paul Gawaz. The funder had no role in study design, data collection, data analysis, data interpretation, or writing of the manuscript.

Data availability statement
The data that support the findings of this study are available from the corresponding author (KALM) upon reasonable request. The data are not publicly available due to the sensitive nature of the data collected for this study.