key: cord-0930757-reuqypc0
authors: Miggiolaro, Anna Flavia Ribeiro dos Santos; Motta Junior, Jarbas da Silva; de Paula, Caroline Busatta Vaz; Nagashima, Seigo; Malaquias, Mineia Alessandra Scaranello; Carstens, Lucas Baena; Novais, Andrea Moreno; Baena, Cristina Pellegrino; de Noronha, Lucia
title: Covid-19 cytokine storm in pulmonary tissue: anatomopathological and immunohistochemical findings
date: 2020-11-12
journal: Respir Med Case Rep
DOI: 10.1016/j.rmcr.2020.101292
sha: 85e359626cbe4288374e9551a6e4b07494cb133d
doc_id: 930757
cord_uid: reuqypc0

The COVID-19 pandemic is a worldwide threat, and information on physiopathological aspects of the disease is limited. Despite efforts in searching treatment options, a better understanding of the SARS-CoV-2 pathways can contribute to managing severe cases. In this study, we aim to describe pathological and immunopathogenic findings of two different cases, both in the high-risk group. Post-mortem lung biopsies were analyzed by traditional and immunohistochemical methods. Tissue expression of innate and adaptive immune response biomarkers was tested. We observed a higher innate response in case 1 with an abundance of mast cells, scarce CD8+ lymphocytes, high expression of TNF-alpha, and almost absent adaptative immune response. In case 2, the adaptative immune response was present, with numerous CD8+ lymphocytes and higher levels of IL-4 and TGF-beta. Both cases converged to a prothrombotic state expressing high IL-6, followed by ICAM-1 expression and endotheliites leading to systemic inflammatory response syndrome. In conclusion, differences in age and comorbidities and immune response described here may be related to the SARS-CoV-2 delay in the adaptative immune response, evolution stage of diffuse alveolar damage, and progression for systemic inflammatory response syndrome.

The world is currently going through the COVID-19 pandemics, and the death 2 toll is still on the rise worldwide. Amidst this troubling time, researchers and health 3 care providers strive to find the best therapeutic strategies, and despite all efforts, 4 there is still no robust evidence towards a definitive treatment. 5

Patients infected with the SARS-CoV-2 tend to develop acute respiratory 6 distress syndrome (ARDS) presenting ground-glass opacities on chest computed 7 tomography exams [1] , and studies show about 3.2% of patients in China required the 8 onset of assisted breathing therapies such as intubation [2] . 9

The SARS-CoV-2 pandemic is unique due to a few notable features in the virus: 10 first, we observe its high affinity to the angiotensin-convertase-enzyme-2 (ACE-2) 11 receptors [3] ; secondly, it is high glycosylated spike proteins that have low 12 immunogenicity, thirdly its recent leap from animal to human meaning that there is a 13 high susceptibility rate. The high viability in the external environment and, finally, 14 delayed adaptive immune response to the virus means patients are vectors for this 15 virus for long periods. All these factors may contribute to the quick and sustained 16 spreading of the virus [4] [5] [6] . 17

Once infection reaches the lungs, pneumocytes and macrophages will initiate 18 an innate immune response favoring a disproportionate immunologic response and 19 endotheliites with microthrombi formation [4] , pathological injuries also recently called 20 endotheliopathy and immunothrombosis [7, 8] . The complement cascade activation is 21 then followed by the phenomenon referred to as a cytokine storm that will culminate 22 on endothelial activation and consequently may cause a Systemic Inflammatory 23

Response Syndrome (SIRS) with multiple organ failure and death [9] . 24

In this article, we described findings of two post-mortem biopsies, the first 25 being an 87-year-old woman with various comorbidities and the second a 53-year-old, 26 obese, man without any other previous illness known. We assessed the immune 27 response to the SARS-CoV-2 infection of these two cases and possible 28 immunopathogenic mechanisms of the therapeutic modalities being tested during this 29 outbreak. The National Research Ethics Committee approves this study (Conselho 2 Nacional de Ética em Pesquisa -CONEP -3.944.734/2020), and families consented to 3 the post-mortem biopsy. 4

Clinical data of both cases were obtained from medical records during 6 hospitalization in the Intensive Care Unit (ICU) at Hospital Marcelino Champagnat in 7 Curitiba-Brazil. Testing for COVID-19 was performed on nasopharyngeal swabs taken 8 during ICU hospitalization, and real-time reverse transcriptase-polymerase chain 9 reaction (rRT-PCR) performed was positive for SARS-CoV2 in both cases. 10

Autopsy procedure: 12

Initially, the Imaging exams such as X-rays and CT's were analyzed to identify 13 the pulmonary segments that have been lesioned, especially the lingular lobe, since it 14 is removed in a more agile and practical technique of minimally invasive biopsy. Once 15 we confirmed a radiographically evident lesion on the lingular segment in the upper 16 left pulmonary lobe, we proceeded to collect the sample through a left anterior mini-17 thoracotomy on the fourth or fifth intercostal space. 18

Sample collection : 19 In cases 1 and 2, the samples were similarly sized (3x3 cm), and the samples 20 were delicately handled and resected using surgical scissors. Both samples presented 21 normal pulmonary consistency. The samples were then inserted into a 10% 22 concentration formalin solution and kept in it for at least 24 hours until blocking and 23 slicing for microscopic analysis 24

The lung formalin-fixed paraffin-embedded (FFPE) samples were stained with 25 hematoxylin and eosin (H&E). The immunohistochemistry technique was used to 26 identify innate and adaptive immunity biomarkers (Figure 1 - J o u r n a l P r e -p r o o f in less than 15 days. The first one, two weeks before the present hospitalization when 1 she had a diagnosis of femur fracture, in which she had contact with a COVID- 19 2 person (assistant physician). The second hospitalization was due to gastrointestinal 3 bleeding (endoscopic diagnosis of Mallory Weiss laceration of the esophageal mucosa). 4

Two days before going to the emergency department, she had flu-like symptoms 5 without a fever. In the first hours of hospitalization, her breathing pattern worsened, 6 requiring mechanical ventilation and vasoactive drugs. The patient's illness progressed 7 with SIRS. 8

Pronation maneuver was necessary to maintain PaO2 / FiO2, but the patient 9 developed multiple organ dysfunction and died on the 8th day of hospitalization. In case 2, areas with preserved alveolar structure interspersed with foci of 1 extensive alveolar fibrosis were observed. In fibrotic areas, we observed extensive 2 alveolar collagenization associated with inflammatory infiltrate (lymphocytes and 3 macrophages) and the presence of thickening of the vascular walls and squamous 4

metaplasia. An intense endothelial activation with signs of endotheliitis was also 5 observed (FIGURE 1). 6

Case 1 revealed a predominance of innate response biomarkers tissue 8 expression, such as ficolin 3 (FCN3) in hyaline membranes, IL-1 and TNF-alpha in 9 pneumocytes and alveolar macrophages and abundant mast cells (CD117). Also, a 10 slight expression of adaptive response tissue biomarkers such as CD8+ lymphocytes 11 and IL-4, IL13, and TGF-beta in type II pneumocytes and alveolar macrophages (FIGURE 12 1 and 2). 13

In case 2, the expression of innate and adaptive response was the opposite of 14 case 1. There was a scarce innate response, such as lower FCN3 IL-1, TNF-alpha CD117 15 tissue expression. Also, a predominance of adaptive response was observed, such as 16 higher CD8, IL-4, IL13, and TGF-beta tissue expression (FIGURE 1 and 2). 17

Markers highly expressed in both cases were CD163 (macrophages), SPHINGO 18 (M2 macrophages), IL6 (pneumocytes and alveolar macrophages), and ICAM-1 19 (activated endothelium and pneumocytes) (FIGURE 1 and 2). 20

In both cases, markers that were equally under-expressed were IL10, IL8, and 21 IL17 in type 2 pneumocytes and macrophages (FIGURE 1 and 2). patients with a multitude of diseases ranging from cases of pneumonia to exposition to 30 toxic fumes [10] . This process is characterized by the typical histological finding of 31 Diffuse Alveolar Damage (DAD) in its acute phase, as well as type II pneumocyte 32 hyperplasia and hyaline membranes, and the condition evolves to tissue organization 1 and fibrosis [11] . In this context, both patients showed evident signs of ARDS in 2 different stages, while patient 1 showed more acute signs of ARDS such as type II 3 pneumocyte hyperplasia and hyaline membranes; patient 2 was already in the 4 organization phase with evident fibrosis and scarce hyaline membranes. 5

Although laboratory findings evidenced SIRS in both cases, there were 6 differences in serum lymphocyte levels at admission and lymphopenia in case 1. There 7 is evidence that lymphopenia is a useful indicator of the severity and hospitalization in 8 . Besides, in case 1, histopathological analysis revealed few 9 TCD8 + lymphocytes in the alveolar septa, which may have been caused by the marked 10 lymphopenia. 11

The D-dimer level was a fourfold increase in case 1, average in case 2, and high 12 levels on admission could effectively predict mortality in patients with . 13 Accordingly, case 1 presented the foci of micro thrombosis in capillaries and small-14 caliber pulmonary vessels. 15

The radiological findings, ventilatory parameters applied, therapeutic 16 administered, clinical and laboratory evidence of impaired tissue oxygenation, and 17 finally, the pronation protocol was comparable in both cases. We may hypothesize 18 that the virus infection per se may cause all pathologic changes into lung-morphology 19 since the ventilation of the patients during the intensive care unit (ICU) hospitalization 20 was gentle and controlled. 21 SIRS caused by Cytokine Release Syndrome (CRS) or Cytokine Storm may have 22 an essential role in the outcome of these two patients [4, 9] . Diffuse Alveolar Damage 23 (DAD) is a sign in critically ill COVID-19 infected patients [14, 15] , and, during its 24 activation, the inflammatory response must be strictly regulated to prevent systemic 25 damage (SIRS). 26

The lung injury caused by SARS-CoV-2 appears to begin in the upper airways 27

where the virus replicates in the ACE-2 receptor rich respiratory mucosa [5, 6] . After 28 reaching the alveoli, the virus begins to replicate in the pneumocytes, causing cell 29 death and alveolar septum collagen exposure. The pneumocyte necrosis causes IL-6 30 and TNF-alpha to be released and mast cell degranulation [4] . This scenario would lead 31 to the triggering of the coagulation and complement cascades; it would also increase 32 vascular permeability [9] . A sequent formation of hyaline membranes characteristic of 1 the acute phase of DAD (case 1) is seen. In approximately two weeks of evolution, DAD 2 initiates its resolution phase, and the hyaline membranes are replaced by type II 3 pneumocyte hyperplasia, a chronic inflammatory process followed by pulmonary 4 alveolar fibrosis, as shown in case 2 [14, 15] . Notably, in case 1 (87-year-old woman), 5

there is no significant alveolar edema, commonly seen in acute cases of DAD. 6

Neutrophils and lymphocytes in the alveolar septa are also not seen, another 7 characteristic commonly seen in viral interstitial pneumonitis. Neither case showed 8 signs of secondary bacterial infections, a common cause of death for patients with viral 9 pneumonitis and DAD [14] . 10

Our combined results showed an increased number of mast cells (CD177) with 11 TNF-alpha, IL-6, and IL-1 release (Th2 response) in case 1 and lower levels of these 12 parameters in case 2. We also observed a high number of macrophages (CD163) in 13 both cases. Mast cells and macrophages releasing TNF-alpha in airways with a Th2 14 response predisposition, as observed in case 1, have high potential to contribute to 15 DAD [16, 17] . However, in case 2, the Th2 response was still found. In our observation, 16

case 1 showed few Th1 immune responses since the number of TCD8+ lymphocytes 17 was discrete, contrary to case 2. Different stages of lung injury among cases could 18 explain this discrepancy. Case 1 might have died before the fibrotic response was 19 activated, and the immune senescence might play a role as well. While there is a four-20 day gap between both patients from hospitalization date to the death date, the lack of 21 fibrotic response and inflammatory infiltrate might not be only to the time of disease 22 progression but also to the patient's ability to respond to the infection and the 23 effective immune response. Thus, the age gap and inherent predisposition to a Th1 or 24

Th2 response may play significant roles in the final lung fibrosis and tissue remodeling. 25

The spectrum of SARS-CoV-2 injuries and their symptoms, from asymptomatic 26 to the most severe forms, could also be related to the individual genetic background 27 that could drive the initial prevalent Th2 response without an effective antivirus Th1 28 immunity since this Th1 response could be activated later. Another explanation could 29 be the high spike proteins glycosylation present in SARS-CoV-2, which would hinder An unusual lack of alveolar neutrophils was found and could be explained by 3 the low expression of IL-8 and IL-17 in both cases. As the IL-17A/IL-8 are pro-4 inflammatory cytokines mainly dependent on activated T cells, the poor (case 1) and 5 later (case 2) Th1 response of cases could explain this phenomenon [19] . 6

Deposition of Ficolin-3 (FCN3) within lung septal microvasculature has been 7 described, promoting complement system activation in COVID-19. We observed the 8 expression of FCN3 in both cases, more evidently in case 1 (acute phase). The 9 complement cascade precedes endothelial injury and leads to the clotting pathway 10 activation culminating in systemic micro thrombosis and multiple organ failure [20] . 11

Moreover, endotheliopathy and immunothrombosis may have a crucial role in 12 SIRS observed in patients with COVID-19, as some studies described the endothelium 13 as a potential target of the SARS-CoV-2 since they are described to express ACE-2 type 14 receptors [21] . We also observed high expression of ICAM-1, suggesting endothelial 15 activation as a common finding in both cases. 16 IL-6 levels have been associated with mortality and could be an indicator of its 17 protagonism in the cytokine storm in COVID-19 [22] , as observed in both our cases. 18

Regarding multiple organ failure, the patient's advanced age in case 1 could 19 compromise the immune system activation "per se," as shown by the inadequate 20 adaptive immune response observed with scarce of TCD8+ lymphocytes, however, 21 even with higher expression of CD8+ lymphocytes in case 2, the fibrosis promoted by 22 (20) In these cases, in response to necrosis of pneumocytes injured by SARS-COV2, we note 39 the massive presence of mast cells (CD117) and macrophages (CD163), capable of 40 secreting cytokines such as TNF-α and interleukins (IL) type 1 and 6, respectively. IL-6 41 and TNF-α could elevate vascular permeability, promoting exudation of plasma 42 proteins, including ficolin 3 (FCN3). FCN3 and other plasmatic proteins inside the 43 alveoli lumen are responsible for promoting opsonization, activating the complement 44 and the coagulation cascade, and, as observed in the present work, promoting hyaline 1 membranes. The presence of IL-1 and TNF-α, inducing massive expression of ICAM-1 in 2 both cases, can promote the activation of endothelial cells and endotheliitis and may 3 be involved in the progression for systemic inflammatory response syndrome (SIRS). 4

APCs can present the antigens for T lymphocytes, via MHC (main histocompatibility  5 complex), and secreting cytokines that play critical roles in the differentiation of T cells 6 into effector cells. APCs express MHC I molecules and present TCD8+ lymphocytes. 7

On the other hand, macrophages and APC activate TCD4+ lymphocytes via MHC II, 8 organizing responses classified into subgroups, such as Th1, Th2, Th17, depending on 9 cell differentiation. In case 1 of the present study, a high immunoexpression of CD117 10 (mast cells) FCN3, TNF-α, IL-1, and Il-6 is observed, suggesting that the inflammatory 11 process is sustained, mainly, through innate immunity. Besides, there is a low number 12 of TCD8+ lymphocytes, suggesting that adaptive immunity is not still adequately 13 stimulated. In contrast, in case 2, in addition to the presence of TCD8+ lymphocytes, 14

there is a marked expression of Sphingosine (M2 macrophages), IL-4, IL1-3, and TGFß, 15

suggesting the presence of adequately stimulated of adaptive immunity with a 16 tendency to resolve with fibrogenesis. A slight immunoexpression IL-8 and IL-17 is also 17 observed, suggesting low stimulation of the Th17 response and explaining the almost 18 absence of alveolar neutrophil migration. 19 20 Figure 2 

Case 1 -Innate Immunity -Photomicrographs from A to D 23

In photomicrograph A it can be observed FCN3 (arrows) expressed along the thick 24 hyaline membranes (asterisk). 25

In B, it observed the expression of IL-6 in the alveolar septum (arrow) by lymphocytes, 26 type II pneumocytes, and alveolar macrophages. 27

In C, it can observe the expression of TNF alpha in the alveolar septum (arrow) by type 28 II pneumocytes and alveolar macrophages. 29

In D, it observed endothelium (arrow) and pneumocytes (asterisk) expressing ICAM-1. 30

Presence of polymorphonuclear cells permeating the vascular wall (dashed arrow) 31 configuring endotheliitis. 32 33

Case 2 -Innate Immunity -Photomicrographs from E to H 34

In photomicrograph A, there is a scarce expression of FCN3 in a few alveoli still with 35 remaining hyaline membranes (arrow). 36

In B, it can observe a marked expression of IL-6 in the areas of fibrosis (arrow). 37

In C, it can observe a discrete expression of TNF alpha (arrow) only in a few alveoli with 38 remaining hyaline membranes (asterisk). 39

In D, it observed a marked expression of ICAM-1 in endothelial cells (arrow) and 40 pneumocytes (asterisk). Vessels show activation of endothelial cells and endotheliitis 41 (dashed arrow). 42 43

Case 1 -Adaptive Immunity -Photomicrographs from A to D 44

In photomicrograph A, it can be observed scattered (about 2 per medium power field) 45 CD8+ T lymphocytes (arrows). Vessel with polymorphonuclear cells on its wall (dashed 46 arrow) configuring endotheliitis. 47

In image B, it observed numerous alveolar macrophages (arrow) expressing 1 SPHINGOSINE, which characterizes the M2 type macrophage. 2

In C, it can be observed a slight expression of TGF-beta in type II pneumocytes (arrow) 3

and alveolar macrophages. 4

The photomicrograph D can observe moderate expression of IL-4 in type II 5 pneumocytes (arrow) and alveolar macrophages. 6 7

Case 2 -Adaptive Immunity -Photomicrographs from E to H 8 In photomicrograph A, it can observe numerous (about 20 to 30 per medium-9 magnification field) CD8+ T lymphocytes (arrow). 10 In image B, it observed numerous alveolar macrophages (arrow) expressing 11 SPHINGOSINE, which characterizes the M2 type macrophage. 12

In C, it observed a moderate expression of TGF-beta in macrophages and fibroblasts 13 (arrow). 14 It can observe moderate to a marked expression of IL-4 (arrow) in macrophages and 15 lymphocytes in photomicrography D. 

Diffuse and bilateral "opacities with groundglass attenuation" Thickening of the pulmonary septum, suggestive of viral pulmonary infection Diffuse and bilateral "opacities with groundglass attenuation", suggestive of viral pulmonary infection

Limited 1 replication of influenza A virus in human mast cells

Hemophagocytic 4 lymphohistiocytosis: a review inspired by the COVID-19 pandemic

Development of a nucleocapsid-based human

coronavirus immunoassay and estimates of individuals exposed to coronavirus 9 in a U.S. metropolitan population

Interleukin-17A activation on bronchial 12 epithelium and basophils: A novel inflammatory mechanism

Complement associated microvascular injury and 16 thrombosis in the pathogenesis of severe COVID-19 infection: A report of five 17 cases

Endothelial 20 cell infection and endotheliitis in COVID-19

C Reactive Protein = 313 mg/L

Total leukocytes = 15,100/neutrophils = 1208 (8%) and lymphocytes = 604 (4%)

Creatinine = 1.1 mg/dL C-Reactive Protein = 146 mg/L

D-dimer= 425 µg/mL (normal)

Total leukocytes = 11,000/ neutrophils = 550 (5%) and lymphocytes = 990 (9%)

Protein = 201 mg/dL

000/ neutrophils = 2400 (16%) and lymphocytes = 1200

Creatinine = 2.65 mg/dL

Protein = 133 mg/dL

/ neutrophils = 3012 (6%) and lymphocytes = 3514 (7%)

Hydroxychloroquine 800 mg/day on the 1st day and 400 mg / day on the other days + Azithromycin 500 mg/day for 5 days

Oseltamivir 150 mg/day for 5 days

Piperacillin + Tazobactam 18 mg/day

Hydroxychloroquine 800 mg/day on the 1st day and 400 mg/day on the other days + Azithromycin 500 mg/day for 5 days

Oseltamivir 150 mg/day for 5 days

Ceftriaxone 2g/day