Abstract
Introduction
On December 2019, a series of cases of pneumonia of unknown origin linked to the Huanan seafood wholesale market in the city of Wuhan, province of Hubei, were reported in China. The causative agent was identified as a novel betacoronavirus, temporarily called 2019-nCov, later classified as SARS-CoV-2. 1 The disease was named by the World Health Organization (WHO) as Coronavirus Disease 2019 (COVID-19). Next-generation sequencing analysis of the viral genome resulted in an 80% sequence identity to SARS-CoV and an even closer relationship to a bat coronavirus, and it was also confirmed that this virus uses the same receptor as SARS-CoV to enter the host cell: angiotensin-converting enzyme II (ACE2).
Coronaviridae is a family of enveloped viruses that includes the genera alpha, beta, gamma and delta. Their natural hosts are mammals and birds, in which they produce different illnesses. The genus alpha comprises common cold coronaviruses that infect humans, and the genus beta include SARS-CoV and MERS-CoV, causal agents of severe acute respiratory syndrome (SARS) and Middle East respiratory syndrome (MERS), respectively, and the newly described SARS-CoV-2. Occasionally, they jump from their natural hosts such as bats, into intermediate hosts – specifically, the civet cat in SARS, the dromedary camel in MERS and, possibly, the Malayan pangolin in SARS-CoV-2, before infecting humans. 2 Coronaviruses are positive-sense, single-stranded RNA viruses, which means that they function as messenger RNA and are translated directly into proteins in the host cells. Their genome of 30 kb is the largest viral RNA. The first two thirds encode replicase genes, while the rest of the genome encodes for structural proteins that comprise the components of the virion – namely, spike (S), nucleocapsid (N), envelope (E) and membrane (N), as well as other non-structural proteins. The S protein is crucial to the infectious process through its direct interaction with the ACE2 receptor in the host cells, and cleavage by the host serine protease TMPRSS2, allowing the virus to enter the cell.3–5 Cells rich in ACE2 receptors are present in the nasal mucosa, alveolar type II cells (ATII) of the lower respiratory tract and cells of other organs – mainly, heart, oesophagus, intestine, kidney and endothelial cells, as has been shown by ACE2 gene expression. 6
The virus is transmitted largely by respiratory droplets from infected individuals to people in situations of close interaction – for example, professional meetings, social, religious, commercial and tourism activities, apart from closer household contacts.7,8
The illness is characterized by a mean incubation period of 5.1 days, with the majority of patients showing symptoms within 12 days of infection. 9 The symptoms first described in a large series of patients from China were fever, headache, sore throat, cough, malaise, with or without mild dyspnoea in 81% of cases, and approximately 20% with gastrointestinal symptoms; while 14% progressed to severe dyspnoea requiring assisted ventilation and 5% develop critical illness with acute respiratory distress syndrome (ARDS) and multiorgan failure (MOF). Chest x-ray (CXR) and computed tomography (CT) scan showed ground glass opacities (GGOs) and consolidations.10,11 Additional mild symptoms such as anosmia and ageusia have been reported recently. 12 The crude mortality rate due to COVID-19 is between 3% and 4%, with great variations depending on age, pre-existing comorbidities and status of the healthcare system in the respective location. 13
As the outbreak spread rapidly to the rest of the world, achieving the status of a pandemic, and healthcare systems became overwhelmed with taking care of the sick and the rising death toll, it became evident that post-mortem (PM) studies would be necessary to aid in understanding this new disease. There was a delay in the publication of a large series of autopsy findings 14 ; however, professional organisms such as the Centers for Disease Control, Royal College of Pathologists and College of American Pathologists issued guidelines for the safe performance of these procedures, as well as recommendations for tissue sampling and molecular testing of these specimens.15,16
The purpose of this review is to summarize, in a narrative and comprehensive manner, the respiratory pathology findings from PM examinations of large autopsy series and unique case reports, perform clinicopathological correlation and hypothesize about possible pathogenetic mechanisms in this new disease.
Pulmonary pathology
Acute lung injury (ALI): Diffuse alveolar damage (DAD), organizing pneumonia (OP) and acute fibrinous and organizing pneumonia (AFOP)
As was expected in a respiratory viral illness, and in light of previous descriptions of histopathologic findings of SARS-CoV and MERS-CoV, many authors have reported the histologic features of DAD at different stages of evolution. 17
The first PM study from Xu et al. 18 was from a 50-year-old man with fever, chills cough, fatigue, shortness of breath and bilateral GGOs on CXR, who was confirmed positive for SARS-CoV-2 by reverse transcriptase polymerase chain reaction (RT-PCR). He developed respiratory failure (RF), required supplemental oxygen and received different treatments including interferon alfa-2b, antivirals, steroids and invasive ventilation, but died on day 14 of illness. Core biopsies of both lungs showed features typical of DAD – namely, oedema, hyaline membranes, desquamation of pneumocytes and presence of multinucleated syncytial cells suggestive of presence of viral cytopathic effect. An interstitial lymphocytic infiltrate was also described. 18
The first autopsy report from the US by Barton et al. reported two patients without pre-mortem diagnosis of COVID-19.
19
One of them showed findings typical of DAD, with thrombi in small pulmonary artery branches and interstitial pneumonia (IP). The decedent was a 77-year-old obese man with history of hypertension (HTN), who presented with fever and chills, and died six days after onset of illness. His PM nasopharyngeal and lung parenchymal swabs were positive for SARS-CoV-2, while a respiratory virus panel was negative. Additionally, immunohistochemical (IHC) studies for lymphocytes showed predominant helper T cells (CD8+) in the interstitium, followed by CD4+ T cells and rare CD20 + B cells. In comparison, the other case did not show DAD, and showed aspiration pneumonia and bronchopneumonia. It was a 42-year-old man with myotonic muscular dystrophy who presented to the emergency room (ER) in critical condition, with fever, shortness of breath and cough, and progressed to sudden cardiac death in a few hours. His nasopharyngeal swabs were positive for SARS-CoV-2, but the PM lung swabs were negative; instead, the lung tissue grew
Tian et al. presented two cases of “early phase” COVID-19 pneumonia that was discovered incidentally in patients undergoing lobectomies for lung cancer. 20 One of the patients was an 84-year-old lady with type 2 diabetes mellitus (DM2) and HTN who acquired COVID-19 during hospitalization for the operation and died on day 29 of hospitalization (day of hospitalization (DH)). On DH6, a chest CT scan showed the tumour and some GGOs. After undergoing surgery (DH12), she presented with wheezing, increased bilateral GGOs, RF and tested positive for SARS-CoV-2. The resected lobe showed oedema, intra-alveolar exudates, fibrin, protein globules, pneumocyte hyperplasia, multinucleated giant cells and mild inflammatory infiltration apart from a lung adenocarcinoma. 20
The most detailed semiquantitative description of histologic features of different stages of DAD was published by Carsana et al. in 38 autopsies of Italian patients who died from COVID-19. 21 The features also included OP, AFOP and IP characterized by lymphoplasmacytic infiltrates. Some patients had focal microcystic honeycombing, indicating progression to fibrosis without evidence of pre-existing chronic fibrosing lung disease. The patients were 33 men and five women, with a mean age of 69 years and many comorbidities including DM2, HTN, past history of malignancy, cardiovascular disease (CVD) and chronic obstructive pulmonary disease (COPD), and had spent a mean of seven days in intensive or sub-intensive care units before demise. All had clinical and radiological findings of IP. 21
Menter et al. published the first comprehensive series of PM studies on 21 predominantly male patients of blood group A, age range 53–96 and one or more comorbidities such as HTN, DM2, CVD, obesity and COPD. 22 They presented with cough, fever, dyspnoea, GGOs, acute on chronic kidney injury and 50% required intubation. All had pre-mortem diagnosis of COVID-19 by RT quantitative polymerase chain reaction (RT-qPCR) from nasopharyngeal swabs or respiratory samples. Length of hospitalization ranged from 0 to 16 days, and all patients died of RF. Histopathologic findings were marked capillary congestion and DAD in both phases: exudative and organizing. The majority showed reactive ATII with multinucleation and microthrombi in alveolar capillaries. 22
Bradley et al. published an autopsy series of 14 patients who died with COVID-19. 23 The median age was 73.5 years (range 42–84) with HTN, DM2, obesity and chronic kidney disease as comorbidities. The majority presented with cough, fever, respiratory distress and showed bilateral opacities on CXR. Twelve of 14 patients had DAD at different stages: hyaline membranes, reactive pneumocytes with multinucleated giant cells and perivascular lymphocytic infiltrates and focal pulmonary microthrombi. No endotheliitis or vasculitis were seen. The IHC studies performed are discussed below. 23
Konopka et al. compared the PM pulmonary findings of three cohorts of patients comprising four community patients with PM diagnosis of COVID-19 who died at home, four inpatients who died in hospital after medical interventions and a control group of eight patients who died of other causes before the SARS-CoV-2 pandemic but required artificial ventilation and had DAD as PM histologic diagnosis. 24 The community patients were predominantly black, with an average age of 55.5 years (range 44–67) and comorbidities such as obesity, CVD, DM2, renal failure and drug abuse.
Inpatients were also predominantly black, with an average age of 56.3 years (range 37–79), days on ventilator ranging from 0 to 16, DM2, asthma, renal transplant, HTN and coronary artery disease (CAD) as comorbidities. Presenting symptoms were fever, cough, myalgia, dyspnoea, headache and gastrointestinal complaints. Characteristics of DAD: hyaline membranes, fibrin thrombi, airspace organization and AFOP-like fibrin were similar between the inpatient, community and control cohorts; however, COVID-19 patients had more focal perivascular inflammation and endotheliitis than control patients. It was concluded that there were no significant morphologic features to differentiate DAD related to SARS-CoV-2 infection from those seen in other causes. 24
Sauter et al. studied eight cases PM and found features of DAD in different phases, and demonstrated that the presence or absence of reactivity for viral proteins by IHC studies depended on the phase of this ALI process. 25
Copin et al. reported PM studies of five patients who died approximately 20 days from presentation of illness. Biopsies (PM) showed AFOP accompanied by hyperplastic/reactive ATII. One of their patients died suddenly on DH5 and showed only interstitial lymphocytic infiltrates and some reactive pneumocytes. 26
Flikweert et al. performed PM biopsies in seven patients (age range 58–83) who died with COVID-19-related ARDS, with 12–36 days from admission to death and a mean time on a ventilator of 20 days. Chest CT showed progressive GGOs. Galactomannan was positive in six patients. Histology showed a predominance of OP, followed by AFOP, DAD and fibrosis, with no evidence of fungal infection. 27
Endotheliitis, microangiopathy, angiogenesis and microthrombosis
Varga et al. described endotheliitis in different organs in three patients, one of whom was a 71-year-old male with CAD, HTN and a renal transplant. He was confirmed positive for SARS-CoV-2 and developed RF requiring mechanical ventilation, which progressed to MOF and demise. His lungs showed evidence of inflammation and damage in the endothelium characterized by inflammatory infiltrates and apoptosis, demonstrated by positive IHC reaction to caspase. 28
In a group of five PM studies with more than two weeks of illness, Copin et al. demonstrated endothelial cell damage characterized by vacuolization and cell detachment in small- and medium-sized pulmonary arteries. 26
Ackermann et al. performed PM examination of lungs from seven patients who died with COVID-19 and compared findings with lungs from seven patients who died with ARDS due to influenza A (H1N1) and 10 uninfected control lungs. 29 They found DAD with interstitial and perivascular infiltration of T lymphocytes as well as pulmonary arteriolar thrombosis of vessels with a diameter of 1–2 mm, with partial luminal obliteration in COVID-19 and H1N1 patients alike; however, there was a higher prevalence of alveolar capillary microthrombi in COVID-19 patients (nine times more). The microvasculature was further examined by scanning electron microscopy and microvascular corrosion casting, demonstrating increased presence of intussusceptive angiogenesis in lungs of COVID-19 patients simultaneously with 69 differentially expressed angiogenesis-related genes compared to those with H1N1. Transmission electron microscopy (TEM) showed evidence of endothelial damage. 29
Magro et al. reported five patients with RF and purpuric skin lesions due to severe COVID-19. 30 Two of the patients demised and had limited PM exams. One of them was a 62-year-old male with DM2, CAD, heart failure, previous hepatitis C and end-stage renal failure, who presented to the ER with severe hypoxemia, HTN, bilateral opacities in the lungs on CXR and died a few hours later. PM examination of the lungs showed pulmonary haemorrhage with endothelial cell necrosis and significant fibrin deposition within septal capillaries and exudation of neutrophils into alveolar spaces. No features of DAD or viral cytopathic effect were seen. Immunohistochemistry demonstrated deposits of C4d, C5b-9 and C3d in alveolar septal capillaries They reported similar deposits of C5b-9 within dermal capillaries. The second case was a 73-year-old male with history of smoking, obesity and prediabetes, who presented to the ER with fever, respiratory distress and hypoxemia and required mechanical ventilation. On DH5 he expired after developing MOF. Lung PM examination showed similar features to those described in the first patient; however, in this case, there was some development of features of DAD – namely, hyaline membranes and ATII hyperplasia. This patient also had deposits of C5b-9 and C4d in alveolar capillaries and vessels of a larger diameter, demonstrable by IHC. MASP2 was positive by IHC in the alveolar septa. 30
Fox et al. described 10 autopsies of African American decedents, positive for SARS-CoV-2, aged 44–78 years, who had at least one of the comorbidities of DM2, HTN and obesity and presented with 3–7 days of mild cough and fever that developed into RF or collapse at home. Chest CT scan showed GGOs and features of ARDS. 31
Histologic findings included exudative and proliferative phases of DAD with desquamated ATII with viral cytopathic-like changes, microthrombosis in small capillaries and venules, increased CD61+ megakaryocytes by IHC, increased IHC expression of Von-Willebrand factor in endothelial cells and presence of perivascular infiltrates of CD4+T lymphocytes. 31
IHC evidence of viral proteins in lung tissue
Zhang et al. performed PM transthoracic needle biopsies of a 72-year-old patient with HTN and DM2 who died three weeks after diagnosis of COVID-19. 32 He presented with fever, cough and developed progressive RF requiring assisted ventilation one week after onset of disease. Chest CT scans showed bilateral GGOs. Four cores of lung tissue showed organizing phase of DAD. Fluorescent IHC reaction for viral Rp3 nucleoprotein was localized to the alveolar epithelial cells, including the desquamated ones, and within capillaries of the alveolar wall. 32
Yao et al. reported a case of a 78-year-old woman who was admitted due to a fall and was found to have GGOs and positive PCR test for SARS-CoV-2. 33 Fourteen days later, after three negative nasopharyngeal swabs and resolving lung infiltrates, she was ready to be discharged but died suddenly of cardiac arrest. PM digital PCR and histopathologic studies of biopsies of different organs were performed. PCR was positive only in lung tissue which also showed features of DAD and presence of reactivity with IHC for SARS-CoV-2 Nucleocapsid (N). 33
Magro et al. described the presence of SARS-CoV-2 spike glycoprotein by IHC in the alveolar septa, co-localized with C3d and C5b-9 in one of the patients who succumbed to COVID-19. 30
Sauter et al. described a group of eight patients with equal number of males and females, median age 57.5 (range 39–65) with obesity and comorbidities: DM2, HTN, CVD, hyperlipidaemia and COPD. 25 They presented with fever, cough, respiratory distress and bilateral GGOs, and one was dead on arrival. The lungs showed DAD in exudative and organising phases, platelet and fibrin microthrombi, and thrombi in larger-calibre vessels. IHC studies with monoclonal antibodies against SARS-CoV viral nucleoprotein and S2 spike protein were found reactive only in cases of exudative phase of DAD, but not in those of proliferative phase. Reactivity was localized to hyaline membranes and reactive/regenerating ATII, as well as some endothelial cells and venules and in one case in bronchial epithelial cells. 25
Bradley et al. reported some reactivity in desquamated alveolar cells and macrophages to SARS-CoV-2 spike protein by IHC. 23
TEM evidence of “viral particles” in lung tissue
Yao et al. described “coronavirus particles” measuring 70–100 nm present in bronchial epithelial and ATII cells. 33
Carsana et al. described particles suggestive of virus in 9/10 from a total of 38 autopsy cases. The viral structures had a mean diameter of 82 nm and were found within cytoplasmic vacuoles in type I alveolar pneumocyte (ATI), ATII and in macrophages. No viral structures were found within multinucleated giant cells or endothelial cells in the lungs. 21
Ackermann et al. reported the presence of “SARS-CoV-2” within endothelial cells and in the extracellular space. 29
Bradley et al. reported the presence of “viral” particles measuring 7–100 nm with spike-like structures in tracheal epithelial cells, ATI and ATII, and in sloughed cells in the alveolar space. 23
Pulmonary embolism (PE)
Wichmann et al. performed a prospective cohort study of 12 consecutive autopsies of COVID-19 patients who died in hospital, mainly of RF, while some had sudden cardiac death. 34 The median age was 73 years (range 52–87), and the majority were male and had comorbidities such as CAD, COPD and asthma. PM CT showed bilateral consolidations. Seven of 12 patients had deep vein thrombosis (DVT) that was unsuspected before death, and, in four cases, PE was the cause of death. The majority also had different stages of DAD. 34
Edler et al. presented the first consecutive series of 80 full autopsies of decedents with confirmed COVID-19 infection, with complete macroscopic study, but histopathology limited to 12 cases. The age range was 52–96 (median 82.4), predominantly males, with comorbidities such as obesity, CAD, lung disease, kidney disease and DM2. DVT was present in 40% of cases and PE was found in nine out of 80 cases. Eight of 12 cases studied histologically showed DAD in different stages of evolution. 35
Bronchiolitis and tracheitis
Lymphoplasmacytic infiltrates of the airways, including pharynx, trachea, bronchi and bronchiole have been described in several studies.19,22,34
Secondary infections: bronchopneumonia
In the series by Menter et al., 50% of the cases showed secondary bronchopneumonia superimposed on DAD. 22 Wichmann et al. reported additional purulent bronchitis. 34
Discussion
Soon after the time of discovery of SARS-Cov-2, it was learnt that the entry site into the host cell is the ACE2 receptor in conjunction with the serine protease TMPRSS2. 5 Cells rich in ACE2, such as those of the oral mucosa and nose, are the port of entry of the virus, from which the virus progresses to the lower respiratory tract, gastrointestinal tract, kidney and other organs rich in these receptors. 36 This explains the ageusia and, in part, the anosmia described as initial symptoms. As the virus invades the host cells and unleashes its replicative machinery, it produces injury and death as part of its replication process. 37 This is manifested in morphologic studies of the tissues. Ackermann et al. showed presence of “viral structures” within damaged endothelial cells in the lung by TEM. 29 Release of the virus causes the cell to undergo pyroptosis producing danger-related molecules – namely, adenosine triphosphate and nucleic acids – which are detected by adjacent cells and macrophages, starting the production of pro-inflammatory cytokines and chemokines that attract monocytes, macrophages and T cells to the site of infection. This explains the presence of T lymphocytes in the sole finding of pneumonitis without evidence of DAD in a patient from a study by Copin et al. who died on DH5. This patient was at the very beginning of the inflammatory process, and was categorized as a probable clinical L phenotype (low elastance), based on Gattinoni et al. dual phenotype hypothesis. 38 If the immune system is healthy and the viral inoculum is probably not too large, 39 CD8+T cells recognize infected cells and eliminate them, CD4+ cells mediate immune response, antibodies bind and inactivate virus, while macrophages phagocytose apoptotic cells and neutralized virus, restoring the cell to a normal state in 80% of cases infected by the virus. In remaining patients, however, the immune response is dysfunctional and gives rise to excessive infiltration of monocytes, macrophages and T cells, systemic cytokine storm, pulmonary oedema and pneumonia, as well as widespread inflammation and multiorgan damage and MOF. 37 Recruiting of immune cells into the airspaces, including neutrophils mediates secretion of proteases, reactive oxygen species and neutrophils extracellular traps that cause epithelial and endothelial injury. An example of this manifestation of the inflammatory process was described by Barnes et al. 40
Concomitantly, the binding of SARS-CoV-2 to ACE2 receptors results in the downregulation of ACE2, with an increase in oxidative stress and inflammation that results in endothelial cell dysfunction and an increase in the Von-Willebrand factor and the development of a procoagulant state with microthrombosis, angiogenesis and direct damage to ATII. 41 The inflammatory process, coupled with an inadequate clearance of the virus from infected cells, gives rise to DAD with its typical histologic findings – difficulty in gas exchange, hypoxemia and shortness of breath – which may progress to RF. The majority of cases reported in a series of autopsies of COVID-19 showed features of DAD at different stages of evolution in the cases described previously,18–23,25 and was described as 75% of total cases in a recent systematic review. 42 However, as pointed out by Konopka et al., the features are indistinguishable from non-SARS-CoV-2 causes of ALI. This is to be expected, since DAD, OP and AFOP are stereotypical patterns of reaction of the lung tissue to injury of different origins, both exogenous and endogenous. 24 A way to prove direct causation of the pulmonary lesion by the virus is to demonstrate its presence in the tissue, as we reported previously.23,25,30,32,33 One group made an interesting observation of IHC evidence of viral proteins in the lungs only during the exudative phase and not the proliferative phase of DAD. 24 This may indicate that, at least in some cases, continuing injury of the tissues is not driven by viral replication anymore, but probably by the unleashed cytokine storm. 37 This finding may also have implications related to case management and isolation to limit cross-transmission; however, the only way to determine infectivity is through plaque assays, only available at biosafety level 3 and not amenable to daily clinical use. 43 Many groups also demonstrated to presence of “viral particles” of sizes and appearances suggestive of coronavirus.21,23,29,33 It has been pointed out, however, that TEM interpretation of viral structures requires expertise, as many normal structures can be misidentified for viruses, mainly rough endoplasmic reticulum with its attached ribosomes, creating the illusion of “spikes”, and TEM may need to be aided by immunogold localization of the viral proteins to make a solid identification of these structures.44,45 Finding predominant endotheliitis and microthrombotic injury in the absence of DAD features in specific cases is explained by the direct viral invasion of endothelial cells and has been stated as a possible explanation of “atypical ARDS” in COVID-19 patients. This is characterized by hypoxemia, with a high compliance and low ventilation/perfusion ratio and low elastance, referred to as the L phenotype.29,46 Recent investigations, however, have failed to confirm the hypothesized two phenotypes and this is a field of continuous research in conjunction with clinicopathological correlation. 47 Endothelial injury accompanied by an increased Von-Willebrand factor and inflammation-driven prothrombic environment is associated with diffuse organ microthrombosis. Intussusceptive angiogenesis was described in COVID-19 infection; however, it is not specific, and may represent a situation of emergent need of re-establishment of oxygen supply after severe endothelial damage. The fact that it was not seen in the H1N1 control group may be due to different clinical circumstances, such as lack of mechanical ventilation in the COVID-19 group. 29
The contributions of pathology and forensic sciences during this pandemic have been paramount in order to determine an accurate cause of death and shed light on the pathogenesis of the disease, guide research, treatment and new discoveries.
