postmortem findings associated with sars-cov-2

Dow

nloadedfrom

httpjournalslwwcom

ajspby

BhDMf5ePH

KbH4TTIm

qenVA+lpWIIBvonhQ

l60Etgtdnn9T1vLQWJq+R

2O4Kjt58

on01232021

DownloadedfromhttpjournalslwwcomajspbyBhDMf5ePHKbH4TTImqenVA+lpWIIBvonhQl60Etgtdnn9T1vLQWJq+R2O4Kjt58on01232021

Postmortem Findings Associated With SARS-CoV-2Systematic Review and Meta-analysis

Swati Satturwar MD Mary Fowkes MD PhDdagger Carol Farver MDDagger Allecia M Wilson MDDaggerAlbino Eccher MDsect Ilaria Girolami MDsect Elisabet Pujadas MD PhDdagger Clare Bryce MDdagger

Fadi Salem MDdagger Siraj M El Jamal MDdagger Alberto Paniz-Mondolfi MD PhDdaggerBruce Petersen MDdagger Ronald E Gordon PhDdagger Jason Reidy MScdagger Filippo Fraggetta MD∥

Desiree A Marshall MDpara and Liron Pantanowitz MD MHADagger

Abstract Coronavirus Disease 2019 (COVID-19) caused by thenovel Severe Acute Respiratory Syndromendashassociated Coronavirus2 (SARS-CoV-2) has become a global threat to public healthCOVID-19 is more pathogenic and infectious than the prior 2002pandemic caused by SARS-CoV-1 The pathogenesis of certaindisease manifestations in COVID-19 such as diffuse alveolardamage (DAD) are thought to be similar to SARS-CoV-1 How-ever the exact pathogenesis of COVID-19 related deaths remainspoorly understood The aim of this article was to systematicallysummarize the rapidly emerging literature regarding COVID-19autopsies A meta-analysis was also conducted based on data ac-crued from preprint and published articles on COVID-19 (n=241patients) and the results compared with postmortem findings as-sociated with SARS-CoV-1 deaths (n=91 patients) Both autopsygroups included mostly adults of median age 70 years with COVID-19 and 50 years with SARS-CoV-1 Overall prevalence of DADwas more common in SARS-CoV-1 (1000) than COVID-19(809) autopsies (P=0001) Extrapulmonary findings amongboth groups were not statistically significant except for hepatic ne-crosis (P lt0001) splenic necrosis (Plt0006) and while pulp de-pletion (P lt0001) that were more common with SARS-CoV-1Remarkable postmortem findings in association with COVID-19apart from DAD include pulmonary hemorrhage viral cytopathiceffect within pneumocytes thromboembolism brain infarctionendotheliitis acute renal tubular damage white pulp depletion of

the spleen cardiac myocyte necrosis megakaryocyte recruitmentand hemophagocytosis

Key Words autopsy coronavirus COVID-19 pathology SARS-CoV-1

(Am J Surg Pathol 202100000ndash000)

KEY POINTS

(1) A meta-analysis of autopsy findings was conductedbased on data from articles on Coronavirus Disease2019 (COVID-19) (n= 241 patients) and Severe AcuteRespiratory Syndromendashassociated Coronavirus 1(SARS-CoV-1) deaths (n= 91 patients)

(2) Diffuse alveolar damage (DAD) was the mostcommon cause of mortality in both viral groups

(3) Notable extrapulmonary postmortem findings associatedwith COVID-19 include thromboembolism brain in-farction endotheliitis acute renal tubular damage whitepulp depletion of the spleen cardiac myocyte necrosismegakaryocyte recruitment and hemophagocytosis

INTRODUCTIONAutopsies remain the gold standard to understand the

pathogenesis of new and emerging diseases Autopsies per-formed during previous coronavirus pandemics due to SevereAcute Respiratory Syndrome (SARS) caused by SARS-CoV-1 in 2002 and Middle East Respiratory Syndrome(MERS) caused by MERS-related coronavirus (MERS-CoV)in 2012 provided insights into their pathogenesis that con-tributed to improved patient management12 Since the initial2019 outbreak in Wuhan China almost every country hassubsequently been affected by COVID-19 caused by the novelSevere Acute Respiratory Distress Syndrome Coronavirus 2(SARS-CoV-2) While most infected individuals have recov-ered from COVID-19 many have become seriously ill espe-cially the elderly and those with underlying comorbidities (eghypertension diabetes)3 Unfortunately the mortality rateamong severely ill patients despite medical intervention suchas ventilation has been high with an overall global mortalityrate of 7 to 1545

From the Department of Pathology University of Pittsburgh MedicalCenter Pittsburgh PA daggerDepartment of Pathology Icahn School ofMedicine at Mount Sinai New York NY DaggerDepartment of Pathol-ogy University of Michigan Ann Arbor MI sectDepartment ofPathology and Diagnostics University and Hospital Trust of VeronaVerona ∥Department of Pathology Cannizzaro Hospital CataniaItaly and paraDepartment of Pathology University of WashingtonMedical Center Seattle WA

Conflicts of Interest and Source of Funding The authors have disclosedthat they have no significant relationships with or financial interestin any commercial companies pertaining to this article

Correspondence Liron Pantanowitz MD MHA Department of Path-ology University of Michigan NCRC Building 35 2800 PlymouthRoad Ann Arbor MI 48109-2800 (e-mail lironpmedumichedu)

Supplemental Digital Content is available for this article Direct URLcitations appear in the printed text and are provided in the HTMLand PDF versions of this article on the journalrsquos website wwwajspcom

Copyright copy 2021 Wolters Kluwer Health Inc All rights reserved

ORIGINAL ARTICLE

Am J Surg Pathol Volume 00 Number 00 rsquorsquo 2021 wwwajspcom | 1

Copyright r 2021 Wolters Kluwer Health Inc Unauthorized reproduction of this article is prohibited

Studies have shown that SARS-CoV-2 is an RNAvirus that belongs to the same family of beta coronavirusesas SARS-CoV-1 and MERS-CoV Minor differences inthe spike protein of SARS-CoV-2 are associated with itsincreased transmission rate6 Both SARS-CoV-1 andSARS-CoV-2 use the angiotensin-converting enzyme 2(ACE-2) receptor for binding to host cells It is speculatedthat the pathogenesis of COVID-19 is similar to that ofSARS Although acute respiratory distress syndrome(ARDS) and its histologic hallmark DAD is consideredto be the most common cause of death in COVID-19 thepathologic basis of why ARDS occurs only in a subset ofpatients and why only a subset of these afflicted in-dividuals die is unclear78 Furthermore the clinical courseof severe and fatally ill patients with COVID-19 has beencomplicated by cardiac injury shock and thrombosis in-cluding pulmonary embolism and stroke9

The literature regarding COVID-19 is rapidlyemerging with articles progressively unraveling thepathogenesis and underlying pathology of this infectiousdisease based largely on either gross examination onlylimited autopsy case reports small postmortem series andpartial autopsy findings10ndash14 Despite the active appeal toperform autopsies on COVID-19 patients the currentoverall autopsy rate appears to be low15ndash18 The paucity ofautopsy data is not surprising because initially there wererecommendations to suspend postmortems on patientswith suspectedconfirmed COVID-19 infection19 Alsoperforming postmortem examinations on COVID-19 pa-tients is risky and requires an autopsy facility with ap-propriate biosafety accessories20 Fortunately there hasbeen a recent drive to boost postmortem services duringthis pandemic21 This is important because many criticalquestions about COVID-19 remain to be answered Forexample apart from ARDS what role do other processessuch as direct nonpulmonary tissue viral infection im-mune dysregulation primary thrombosis disseminatedintravascular coagulation (DIC) and secondary infectionsplay in fatal COVID-19 disease22ndash25 Clearly there is anurgent need to characterize the underlying pathology ofCOVID-19 to better understand the pathogenesis of thisdisease and more effectively treat infected patients

The aim of this review article was to systematicallysummarize the available emerging literature regardingCOVID-19 autopsies The intent was to additionallyconduct a meta-analysis to compare postmortem findingsin decedents of COVID-19 to previously published find-ings in SARS-CoV-1 associated deaths

MATERIALS AND METHODSA review of the English and non-English literature in

online archival databases up to June 2020 was performedThere were 15 articles about autopsies on patients withconfirmed SARS-CoV-1126ndash39 and 24 articles detailingautopsies on patients with confirmed COVID-1911ndash1440ndash59

that were included (Supplemental Digital Content Table 1httplinkslwwcomPASB82) Two of the SARS-CoV-1articles referred to the same group of 8 patients2627 but

because they provided different clinical and pathologicdetails in each paper both were included without dupli-cating the patient count A total of 332 patients wereidentified including 91 with SARS-CoV-1 infection and241 with COVID-19 All of the papers concerning SARS-CoV-1 were previously published whereas only 15(625) of the papers about COVID-19 autopsies wereaccepted in the peer-reviewed literature at the time ofsubmitting this review with the remaining preprint articlesanalyzed available on medRxiv or bioRxiv One COVID-19 paper of a pregnant woman with COVID-19 has sincebeen withdrawn by the authors47

Specific patient age was provided for 131 patients(60 with SARS-CoV-1 71 with COVID-19) allowing foran age comparison for both patient groups The normalityof the distributions of patient age was examined using theShapiro-Wilk normality test As the data were not nor-mally distributed the Mann-Whitney test was used tocompare SARS-CoV-1 and COVID-19 patient ages Onearticle provided age ranges for each of the 11 patients intheir study52 and these were added to a categorical sum-mary of age with the other 131 patients The Pearson χ2test was used to determine if the proportion of cases withSARS-CoV-1 or COVID-19 was different among catego-rical data variables (sex age category raceethnicitydocumented comorbidities and microscopic pathologicdetails including inflammation thromboembolic findings andnecrosis) Pooled prevalence and their 95 confidence inter-vals of histopathologic findings was calculated using theDerSimonian-Laird random-effects model after double-arcsine transformation Double-arcsine transformation is theappropriate choice when the sample size is small andor ex-treme proportions need to be handled6061 A formal Egger testdetected no publication bias62 Statistical significance was as-sumed at P-value lt005 Analyses were performed using IBMSPSS Statistics 22 and software R 362 (R Foundation forStatistical Computing Vienna Austria) with R Studio125033 environment (R Studio Inc Boston MA)

The selected images included in this manuscript wereacquired from a total of 157 autopsies performed at theaforementioned institutions as well as from 26 personalconsultation cases of autopsies performed at other in-stitutions on patients with confirmed SARS-CoV-2

Autopsy ProcedureSince the outbreak of COVID-19 several articles have

emphasized the need to conduct autopsies using standardizedprotocols and recommended using at least a Biosafety Level-2 autopsy facility2063ndash65 Guidelines were also proposed forforensic pathology66 Although performing a complete bodyautopsy is ideal alternative options utilized to minimize therisk of SARS-CoV-2 transmission to autopsy staff have in-cluded performing limited postmortems with in situ samplingor minimally invasive puncturecore biopsy autopsies111267

of only major target organs with or without opening theentire body Autopsy procedure recommendations have in-cluded utilizing adequately trained staff minimizingthe number of people in the autopsy suite and allowing onlyone person to handle knives The autopsy facility should

Satturwar et al Am J Surg Pathol Volume 00 Number 00 rsquorsquo 2021

2 | wwwajspcom Copyright copy 2021 Wolters Kluwer Health Inc All rights reserved


include an airborne infection isolation room with negativepressure and a minimum number of air changes per hour Ifan airborne infection isolation room is not available aportable high-efficiency particulate air recirculation unit hasbeen recommended to reduce aerosols

In addition to universal precautions additional air-borne safeguards as well as eye protection are recom-mended during these autopsies Personal protectiveequipment recommendations include double surgicalgloves with cutproof synthetic mesh gloves in betweenimpermeablefluid-resistant gown waterproof apron faceshield or goggles N-95 or higher disposable respirators aswell as donning surgical scrubs shoe covers and a surgicalcap Postmortem swabs for COVID-19 testing have beenrecommended from the upper respiratory tract (eg naso-pharyngeal swab) or lower respiratory tract (eg swabfrom each lung)6869 Aerosol-generating procedures(eg use of an oscillating bone saw) are to be avoided Ifbrain removal is performed using an oscillating saw theuse of a vacuum shroud or containment box has beenrecommended70 Standard precautions along with the useof personal protective equipment are also recommendedwhen transferring bodies including disinfecting the out-side of body bags after use Specimens collected for bio-banking have included unfixed tissue swabs or bloodstored at minus80degC as well as formalin-fixed tissue6471

Epidemiology and Clinical SettingArticles about SARS-CoV-1 were mostly from

China (n= 11 73) with 2 published from Singapore and1 each from Taiwan and Canada Articles about COVID-19 arose mostly from the United States (n= 11 46) thenChina (n= 5 21) with 2 published from Switzerlandand 1 each from Austria France Germany Iran Italyand Japan Table 1 summarizes the autopsy proceduresemployed (Supplemental Digital Content Table 2

httplinkslwwcomPASB83) The clinical course (Sup-plemental Digital Content Table 3 httplinkslwwcomPASB84) for COVID-19 in these papers ranged from 0 to61 days after the onset of illness to the time of deathcompared with 1 to 108 days for SARS-CoV-1 patientsThe initial clinical presentation for COVID-19 was similarto SARS-CoV-1 The majority of infected patients pre-sented with fever cough shortness of breath myalgia andother variable symptoms depending upon organ involve-ment (eg diarrhea headache)

COVID-19 patients were statistically significantlyolder (median= 700 y) than SARS-CoV-1 patients (me-dian= 500 y U= 94900 Plt 0001) (Table 2) Of noteCOVID-19 patients were more likely than SARS-CoV-1patients to be 60 to 79 years old (χ2= 10383 P= 0001)and 80+ years old (χ2= 12098 P= 0001) Table 3compares patient race among both viral groups showingthat with SARS-CoV-1 infections were restricted to Asianpopulations but that COVID-19 infections present in allexamined ethnic groups With COVID-19 of 234autopsies that specified sex there were 161 (688)males and 73 (312) females With SARS-CoV-1 of 80autopsies that specified sex there were 57 (713) malesand 23 (288) females There was not a significantassociation between gender and virus group (χ2= 0168P= 0682) Table 4 and Supplemental Digital ContentTable 3 (httplinkslwwcomPASB84) summarizes pa-tient comorbidities indicating that hypertension was themost common comorbid disease overall Nearly all of thelisted comorbidities were documented in a higher pro-portion for COVID-19 patients with the exception of

TABLE 1 Summary of Type of Autopsy Organs Examinedand Timing of Diagnosis

n ()

Autopsy ParametersCOVID-19(N= 241)

SARS-CoV-1(N= 91)

Autopsy typeWhole body autopsy 130 (539) 57 (626)Partialpunch autopsy 110 (456) 34 (374)Not specified 1 (04) 0 (00)

Organ(s) examinedLung 157 (651) 72 (791)Kidney 124 (515) 31 (341)Liver 92 (382) 44 (484)Heart 100 (415) 30 (330)Spleen 92 (382) 39 (429)Brain 71 (295) 27 (297)Gastrointestinal 53 (220) 28 (308)Other organ(s) 36 (149) 39 (429)Endocrine 26 (108) 24 (264)Genitourinary 15 (62) 19 (209)

Timing of diagnosisAntemortem 236 (979) 88 (967)Postmortem 5 (21) 3 (33)

Days to death (range) 0-61 1-108

TABLE 2 Age Distribution by Virus Group in AutopsyPublications

n ()

Age(y)

COVID-19(N= 82)

SARS-CoV-1dagger(N= 60)

HigherProportion χ2 P

le 39 4 (49) 19 (317) SARS 18318 lt 000140-59 12 (146) 23 (383) SARS 10478 000160-79 41 (500) 14 (233) COVID 10383 000180+ 25 (305) 4 (67) COVID 12098 0001

82 of the 241 (340) COVID-19 patients from included articles had patient-level age information the remaining 86 patients provided only summary in-formation (average range) or no age information

dagger60 of the 91 (659) SARS-CoV-1 patients from included articles had patient-level age information the remaining 31 patients provided summary information(average range) or no age information

TABLE 3 Deceased Patient RaceEthnicity by Virus Groupn ()

RaceEthnicity COVID-19 (N= 157) SARS-CoV-1dagger (N= 67)

Caucasian 44 (280) 0 (00)African American 26 (166) 0 (00)Asian 52 (331) 67 (1000)Hispanic 34 (217) 0 (00)

157 of the 241 (651) COVID patients from included articles provided racedagger67 of the 91 (736) SARS patients from included articles provided race

Am J Surg Pathol Volume 00 Number 00 rsquorsquo 2021 SARS-CoV-2 Autopsy Review and Meta-analysis

Copyright copy 2021 Wolters Kluwer Health Inc All rights reserved wwwajspcom | 3


cancer COVID-19 patients were more likely to have hy-pertension diabetes andor obesity than SARS-CoV-1patients (χ2= 21177 Plt 0001 χ2= 8620 P= 0003χ2= 5433 P= 0020 respectively)

Pulmonary radiology features were similar in bothgroups with a chest x-ray or computed tomography scanfindings showing bilateral ground-glass opacities with orwithout consolidation Laboratory test results were re-ported in some studies The most notable findings forCOVID-19 were leukocytosis with relative lymphocyto-penia variably increased lactate dehydrogenase increasedblood urea nitrogen decreased glomerular filtration rateincreased international normalized ratio and increasedD-dimers11124047485253 SARS-CoV-1 patients were alsoreported to present with lymphocytopenia313638

Pulmonary Pathology FindingsThe gross findings in COVID-19 autopsies were no-

table for heavy lungs with edema congestion with orwithout consolidation and in some cases hemorrhage wasobserved (Fig 1) In COVID-19 cases there was a mildserosanguinous pleural effusion (n= 7) and pleural adhesions(n=8) According to Calabrese et al72 pleural effusions weredetected in more than half of the patients with COVID-19Tracheitis was present in a subset of COVID-19 patients40

The main histopathologic lung findings in availablestudies are summarized in Table 5 (Supplemental DigitalContent Table 4 httplinkslwwcomPASB85 and Supple-mental Digital Content Table 5 httplinkslwwcomPASB86) The spectrum of microscopic abnormalities seen inCOVID-19 overlapped with those reported in SARS-CoV-1All of the SARS-CoV-1 patients showed DAD which wasstatistically significantly higher than COVID-19 (χ2=12029P=0001) The majority of COVID-19 patients exhibitedfeatures of DAD (Fig 2) with most cases showing early acuteDAD (Fig 3) Histologic features suggestive of the differentphases of DAD reported in COVID-19 autopsies includedacute DAD (n=74) proliferative DAD (n=22) mixed acuteproliferative DAD (n=42) and fibroticchronic DAD(n=20) For SARS-CoV-1 autopsies the histologic featuresof the different phases of DAD included acute DAD (n=39)(Fig 4) proliferative DAD (n=20) mixed acuteproliferativeDAD (n=23) and fibroticchronic DAD (n=33) Additionalfindings included severe type II pneumocyte hyperplasia

desquamation of pneumocytes squamous metaplasia focalfibroblast plugs interstitial thickening mild patchy chronicinflammation intra-alveolar hemorrhage (Fig 5) mega-karyocytes144346 and hemophagocytosis55 A significantlyhigher proportion of SARS-CoV-1 patients had hemorrhageandor vascular injury identified in their lungs (χ2=28643Plt0001 χ2=14681 Plt0001 respectively) Vascularinjury was noted in both viral groups with most of theinvolved COVID-19 cases showing lymphocytic endo-theliitis44475153 with41 or without necrosis of the vessel wallcytoplasmic vacuolization or endothelial detachment56 Vasc-ular injury described in the articles was mainly endotheliitisdue to inflammation of small capillary sized blood vesselscapillaritis Very few articles included the term vasculitis The

TABLE 4 Documented Patient Comorbidities by Virus Groupn ()

Comorbid Disease COVID-19 (N= 234) SARS-CoV-1dagger (N= 35) Higher Proportion χ2 P

Hypertension 131 (560) 5 (143) COVID 21177 lt 0001Diabetes 68 (291) 2 (57) COVID 8620 0003Obesity 32 (137) 0 (00) COVID 5433 0020Lung disease 31 (132) 1 (29) COVID 3136 0077Heart disease 83 (355) 8 (229) COVID 2164 0141Kidney disease 31 (132) 1 (29) COVID 3136 0077Cancer 25 (107) 4 (114) SARS 0018 0895Immunocompromised 12 (51) 0 (00) COVID 1879 0170

234 of the 241 (971) COVID-19 patients from included articles had comorbidity details provideddagger35 of the 91 (385) SARS-CoV-1 patients from included articles had comorbidity details provided

FIGURE 1 Right lung shown in the sagittal section from a68-year-old woman who died of COVID-19 after 3 weeks ofventilation for ARDS showing diffuse hemorrhagic congestion(lung weight=1130 g normal adult mean weight=300 to350 g)




literature was inconsistent in providing details about vascularinjury and ranged from endothelial denudation to vasculitisIn COVID-19 autopsies for the reported cases with smallvesselcapillary endotheliitis (n=28) most were due to lym-

TABLE 5 Key Microscopic Lung Pathology FindingsCOVID-19 SARS-CoV-1

Histopathology Total n () Total n () Higher Proportion χ2 P

DAD 157 127 (809) 54 54 (1000) SARS 12029 0001Hyaline membranes 157 126 (803) 54 49 (907) SARS 3122 0077Inflammatory cellsdagger 157 115 (732) 54 45 (833) SARS 2487 0135Microthrombi 157 93 (592) 54 33 (611) SARS 0059 0808Hemorrhage 157 56 (357) 54 42 (778) SARS 28643 lt 0001Vascular injuryDagger 157 33 (210) 54 26 (481) SARS 14681 lt 0001Pulmonary embolism 157 13 (83) 54 6 (111) SARS 0393 0531

Approximate number of different phases of DAD as per the histopathologic data provided in the articles For COVID-19 acute DAD (n= 74) proliferative DAD(n= 22) mixed acuteproliferative DAD (n= 42) and fibroticchronic DAD (n= 20) For SARS-CoV-1 acute DAD (n= 39) proliferative DAD (n= 20) mixed acuteproliferative DAD (n= 23) and fibroticchronic DAD (n= 33)

daggerInflammatory cells were present in the interstitium predominantly For COVID-19 lymphocytes (n= 76) plasma cells (n= 6) neutrophils (n= 3) macrophages (n= 3)and inflammatory cell type not specified (n= 1) For SARS-CoV-1 macrophages (n= 34) and lymphocytes (n= 25)

DaggerVascular injury described in the articles was mainly endotheliitis due to inflammation of small capillary sized blood vesselscapillaritis Very few articles included theterm vasculitis Literature was not consistent in providing details about vasculitis and included endothelial denudation also as vasculitis For COVID-19 small vesselcapillary endotheliitis (n= 28) vascular injury withdue to denudation of small to medium-sized vessel (n= 5) For SARS-CoV-1 small vessel endotheliitis (n= 9) andvascular injury withdue to endothelial denudation (n= 17)

FIGURE 2 Forest plot for the prevalence of DAD in reportedautopsy studies with COVID-19 versus SARS-CoV-1 (SARS)as the main subgrouping variable CI indicates confidenceinterval

FIGURE 3 DAD associated with COVID-19 Hyaline mem-branes are present throughout the alveolar spaces in this acutephase of DAD (hematoxylin and eosin)




phocytic endotheliitis or vascular injury withdue to endo-thelial denudation of small-sized to medium-sized vessels(n=5) In SARS-CoV-1 cases there was also small vesselendotheliitis (n=9) and vascular injury withdue to endothelialdenudation (n=17) Vasculitis described in the papersappeared to be secondary and minimal without true vasc-ular destruction or fibrinoid necrosis Interstitial inflammationwas present but was not a prominent component in COVID-19 In COVID-19 the interstitial inflammatory cells weremostly T lymphocytes (CD4+ CD8+)1143 with fewer Blymphocytes as opposed to increased macrophages seen inSARS-CoV-126 Lymphocytic interstitial pneumonia is not animportant finding in COVID-19 cases as per our literaturereview In general lymphocytic interstitial pneumonia is a veryrare entity and its pathology overlaps considerably with thechronic interstitial pneumonia that can be found in manydifferent interstitial lung disease patterns (eg hypersensitivitypneumonitis and the early phase of acute DAD) Therefore inthe studies where this finding is referenced this histopathologyis most likely referring to a more general chronic interstitialpneumonia that is found in these lungs In a small series of 6COVID-19 patients evaluated by postmortem biopsy Copinand colleagues found lymphocytic pneumonia in 1 patient whodied within a week of presentation The remaining cases in their

study all with longer survival intervals (sim20 d) were diagnosedwith an acute fibrinous organizing pneumonia (AFOP) patternof lung injury56 Acute bronchopneumonia with super-imposed bacterial or fungal infection was present in bothCOVID-19 (n= 29)1114404345465255 and SARS-CoV-1(n= 14) cases126ndash2833

Pulmonary embolism with or without associated deepvenous thrombosis was present in both groups The thrombiin COVID-19 patients were comprised of fibrin131448 andorplatelets4548 and mostly involved small peripheral vessels likecapillaries or arterioles with some cases showing thrombi thatalso affected medium-sized vessels (Fig 6) and the pulmonaryartery In the study by Carsana et al45 both lungs fromCOVID-19 patients showed thrombi occupying gt25 of thelung parenchyma associated with elevated D-dimer levels ingt50 of patients

FIGURE 4 Lung tissue showing diffuse hyaline membraneformation and capillary congestion in early acute DAD from apatient infected with SARS-CoV-1 (hematoxylin and eosin)Note the similarity to Fig 3 Image courtesy of Jeffrey L Myersfrom the University of Michigan Ann Arbor MI

FIGURE 5 Marked alveolar hemorrhage associated with COVID-19 Blood and fibrin expanded alveolar spaces are shown con-sistent with alveolar hemorrhage (hematoxylin and eosin)




Multinucleated giant cells (Fig 7) were common inboth COVID-19 (n= 49) and SARS-CoV-1 (n= 44) casesThese giant cells were shown to be CD68+ macrophages orthyroid transcription factor 1+ pneumocytes in COVID-1911 as opposed to mostly CD68+ macrophages in SARS-CoV-126 Viral-induced cytopathic effect (Fig 8) was noted inseveral COVID-19 (n=30) and some SARS-CoV-1 (n=5)autopsies The viral cytopathic effect described in these papersincluded reactive appearing type II pneumocytes withcytomegaly nuclear enlargement prominent nucleoli andcandidate intranuclear or intracytoplasmic viral inclusionsHowever although SARS-CoV-2 viral particles have beenisolated in both type I and II pneumocytes in severalcomprehensive studies to date73 the giant cells and thosepneumocytes with possible viral cytopathic effect have yet tobe further studied in detail

Several studies have demonstrated direct invasionof respiratory epithelial cells by coronavirus using avariety of ancillary techniques including reversetranscriptasendashpolymerase chain reaction (PCR) trans-mission electron microscopy (TEM) immunofluorescence

in situ hybridization and immunohistochemistry (IHC)directed against the nucleocapsid protein and spike pro-tein In patients with COVID-19 TEM detected the virus intype II pneumocytes more often than in type I pneumocytesbronchial epithelial cells and endothelial cells114045 Involve-ment of type II pneumocytes by SARS-CoV-2 was morecommon than type I pneumocytes as opposed to SARS-CoV-1 that affected more type I pneumocytes1128ndash304045 Schaeferet al74 in their series of 7 autopsies reported that SARS-CoV-2infection of epithelial cells in the lungs and airways was bestdetected during the acute phase of lung injury and wasabsent in the organizing phase of DAD For SARS-CoV-2RNAscopereg probes complementary to the viral spike (S) andthe spike sense strand messenger RNAs permit direct visual-ization of viral transcripts in routine formalin-fixed paraffin-embedded tissue Detection of the spike protein (S) indicatesthe presence of SARS-CoV-2 within the cells while detectionof the anti-sense strand shows the presence of active viralreplication in the infected cells75 For both probes cytoplasmicdot-like hybridization signals (Fig 9) confirm the presence ofnondegraded RNA that increase in numbers proportional to

FIGURE 6 Early organizing thrombi are shown in small (A) medium (B) and large (C) vessels in the lung parenchyma of COVID-19 lungs (hematoxylin and eosin)

FIGURE 7 Multinucleated giant cells (arrow) as shown in thisimage can be detected scattered in alveolar spaces throughoutCOVID-19 lungs (hematoxylin and eosin)

FIGURE 8 Possible COVID-19 viral cytopathic effect (arrow)within pneumocytes exhibiting enlarged pink and smudgynuclei (hematoxylin and eosin) This finding may be seen inDAD of other causes




the viral load in the cells ranging in size variability from finedots to a coalescent of such signals forming larger globulesUltrastructural studies with MERS similarly localized viralparticles within pneumocytes76 Ackermann et al53 alsodetected a greater number of ACE-2 receptors in alveolarpneumocytes lymphocytes and endothelial cells of COVID-19patients compared with other viruses

Cardiac Pathology FindingsGross pathology findings of the heart in patients who

died of COVID-19 included cardiomegaly with ventricularhypertrophy correlating with a clinical history of hyper-tension The main postmortem histopathologic findings in the

heart from available studies are summarized in Table 6 andSupplemental Digital Content Tables 6 and 7 (httplinkslwwcomPASB87 httplinkslwwcomPASB88) There was nota statistically significant relationship between COVID-19 andSARS-CoV-1 for microscopic heart findings Nonspecific de-generative changes were noted in some COVID-19studies111451 Myocyte necrosis in COVID-19 was presentonly as individual cell necrosis111450 focally involving bothventricles similar to SARS-CoV-127 Inflammatory cells inCOVID-19 were comprised of lymphocytes (CD4+gtCD8+)55 focally involving the interstitium and epicardialsurface In COVID-19 there were also cases reported withlymphocytic40 (Fig 10) and eosinophilic myocarditis50 For

FIGURE 9 In situ hybridization for messenger RNA encoding the SARS-CoV-2 spike protein (S) is visualized as dark brown signalslocated in the cytoplasm of infected cells (arrows) low power at times400 (A) and high power at times1000 (B C) with some signalsshowing coalescence into globules (C) The antisense messenger RNA strand of the spike protein is shown (arrows) indicative ofactive viral replication low power at times400 (D)

TABLE 6 Summary of Microscopic Heart Pathology FindingsCOVID-19 SARS-CoV-1


Inflammatory cells 99 24 (242) 8 0 (00) COVID 2500 0114Myocyte necrosis 87 7 (80) 8 2 (250) SARS 2455 0117Acute myocardial infarction 99 3 (30) 8 1 (125) SARS 1845 0174Lymphocytic myocarditis 99 3 (30) 8 0 (00) COVID 0249 0617Eosinophilic myocarditis 99 1 (10) 8 0 (00) COVID 0082 0775




SARS-CoV-1 cases no significant inflammation was notedThrombosis of small vessels was noted in some COVID-19cases (n=5)5155 and infrequently with SARS-CoV-1 cases(n=2)27 Hemophagocytosis was reported in the heart byBryce et al55 Both coronaviruses were detected withinmyocytes and the interstitium by PCR12274046

Genitourinary System Pathology FindingsRenal microscopic findings in both viral groups are

summarized in Table 7 and Supplemental Digital ContentTables 6 and 7 (httplinkslwwcomPASB87 httplinkslwwcomPASB88) There was not a statisticallysignificant relationship between COVID-19 and SARS-CoV-1 with respect to renal findings The main histo-pathologic findings were acute tubular damage (Fig 11)

and focal to diffuse microthrombi in both groupsHypertensive and diabetic changes were noted in a fewstudies that correlated with clinical morbidity404857 In theCOVID-19 group additional postmortem findings reportedincluded the presence of interstitial inflammatory cellscomprised of lymphocytes and macrophages58 and vascularinjury such as endotheliitis57 Other incidental findingsincluded amyloidosis (n=1) and an angiomyolipoma (n=1)A plurality of ancillary tests (TEM IHC immunofluore-scence in situ hybridization) have detected SARS-CoV-2 viralparticles within proximal renal tubules more so than distaltubular epithelial cells40484957 or podocytes4849 (Fig 12)whereas the SARS-CoV-1 virus mostly involved the distaltubular epithelium29 Associated findings with COVID-19autopsies were complement deposition58 and activation ofthrombotic pathways with DIC48 Other genitourinary systemfindings reported in COVID-19 patients included an incidentalrenal oncocytoma prostate vein thrombosis46 benign pro-static hyperplasia damage to the testicular parenchyma77 andtesticular atrophy4346

Brain Pathology FindingsExamination of the brain in COVID-19 patients

showed focal punctate subarachnoid hemorrhages punctatebrain stem hemorrhage and focal hypoxic changes withoutsignificant inflammation or necrosis (Supplemental DigitalContent Table 8 httplinkslwwcomPASB89) A study byBryce et al55 showed thrombi in small vessels with associatedmicrohemorrhages and acute infarction (Fig 13) In 1 casethere was a large cerebral artery infarct55 The majority ofcases exhibited minimal inflammation only slight neuronalloss no myelin loss using Luxol fast blue no microglialnodules and no viral inclusions The study by Gu et al29 onSARS-CoV-1 reported the presence of focal hypoxicchanges SARS-CoV-2 was detected in brain tissue byreverse transcriptase-PCR46 Menter et al48 also detectedslightly higher levels of this virus within the olfactory bulbcompared with the brainstem

Hematopoietic System Pathology FindingsThe bone marrow showed varying degrees of de-

creased trilineage hematopoiesis and in some cases ofCOVID-19 there was left-shifted myelopoiesis1148 (Sup-plemental Digital Content Table 8 httplinkslwwcomPASB89) In addition increased megakaryocytes werereported in association with SARS-CoV-130 In COVID-19 patients there was lymphoid depletion of lymph nodesincreased plasmablasts in lymph nodes and white pulpdepletion of the spleen11405559 One study reported

FIGURE 10 Section of the heart from a patient that died withCOVID-19 with lymphocytic myocarditis showing a diffuselymphocytic infiltrate (hematoxylin and eosin) This patientalso had methicillin-sensitive Staphylococcus aureus and influ-enza A infections Low-level SARS-CoV-2 RNA was detectedwithin the myocardium

TABLE 7 Summary of Microscopic Kidney Pathology FindingsCOVID-19 SARS-CoV-1


Acute tubular damage 121 62 (512) 8 6 (750) SARS 1700 0192Fibrin thrombi 109 7 (64) 8 2 (250) SARS 3623 0057Inflammatory cells 83 6 (72) 8 0 (00) COVID 0619 0431Endotheliitis 109 3 (28) 8 0 (00) COVID 0226 0635




increased immunoblast-like cells in lymph nodes78 IHCidentified specific depletion of CD3+ CD4+ and CD8+

lymphocytes in COVID-1959 All SARS-CoV-1 patients inwhich the spleen was examined also showed white pulpdepletion27ndash29 a finding that was statistically significantlyhigher than in COVID-19 patients (χ2=45377 Plt0001)Necrosis of the spleen was more frequent with SARS-CoV-1(χ2=7511 P=0006) whereas inflammation was morecommonly encountered in COVID-19 (χ2=5195 P=0023)(Table 8) Bryce et al55 reported hemophagocytosis in the bonemarrow lymph node and spleen (Fig 14) Increased CD68+

activated large macrophages were reported in SARS-CoV-1patients29 Thrombi were not documented in any of thespleens examined

Hepatic Pathology FindingsIn both groups of patients studied (Supplemental Digital

Content Tables 6 7 httplinkslwwcomPASB87 httplinkslwwcomPASB88) the liver showed focal hepatocytedegeneration and necrosis a mild portal and periportal lym-phocytic infiltrate rare lobular lymphocytic infiltrate steatosisand lymphocytic endotheliitis A significantly higher pro-portion of SARS-CoV-1 patients had hepatocyte necrosis2737

andor portal inflammation in their liver (χ2=19221Plt0001 χ2=10501 P=0001 respectively) (Table 9) Ahigher proportion of COVID-19 patients had thrombi andorlobular inflammation but there was not a statisticallysignificant difference Also hemophagocytosis was noted byBryce et al55 in their series of patients with COVID-19

FIGURE 11 Renal tubular injury A Section of a kidney from a deceased patient with COVID-19 showing acute tubular damagewith coagulative necrosis of tubules and focal shedding of epithelial cells into the lumen The interstitium shows edema and a fewinfiltrating mononuclear inflammatory cells (hematoxylin and eosin) B Renal tubular injury with vacuolization mitoticfigure (green arrow) and focal interstitial inflammation (hematoxylin and eosin)

FIGURE 12 TEM of SARS-CoV-2 in Vero cells (African green monkey kidney epithelial cells) A Multiple viral particles along the cellsurface are shown displaying a distinct electron-dense surface with radiating peplomeric projections (bar 025 microm) B Ultra-structural details on a replicating vesicle Note that the virions are spherical with some pleomorphism and measuring sim80 to 110nm in diameter The surface is covered by an array of projections (peplomers) with the presence of multiple internal electron-densedots that correspond to cross-sections of the nucleocapsid (bar 02 microm)




Gastrointestinal Tract Pathology FindingsIn a few COVID-19 cases the gastrointestinal tract

showed mesenteric ischemia44 The stomach and intestinesin patients with COVID-19 showed epithelial degenerationnecrosis shedding congestion as well as mild lymphocyte andplasma cell infiltration11 There was focal gastric hemorrh-age40 and in 2 COVID-19 cases small vessel endotheliitis ofthe small intestine was identified44 (Supplemental DigitalContent Table 8 httplinkslwwcomPASB89) The pan-creas was uninvolved in most cases With SARS-CoV-1 pa-tients showed atrophy of their mucosal lymphoid tissue (iedecreased lymphocytes depletion of follicles and a burnt-outappearance of germinal centers) mostly in the pharynx smallintestine and appendix30 Hyaline thrombi in small blood

vessels of the small and large intestines and an associatedpseudomembrane with necrotic mucosa was noted in 1COVID-19 patient46

Endocrine Organ Pathology FindingsIn COVID-19 patients the thyroid gland showed an

incidental goiter in 1 patient43 and the adrenal glands werereported to exhibit nodules in another46 In SARS-CoV-1patients Wei and colleagues studied the thyroid gland indetail and reported the destruction of follicular epithelialcells with their exfoliation into the follicle (SupplementalDigital Content Table 8 httplinkslwwcomPASB89)The presence of apoptosis was confirmed by a Terminaldeoxynucleotidyl transferase dUTP Nick End Labeling

FIGURE 13 Neuropathology findings A Coronal section of the right cerebral hemisphere in a patient that died of COVID-19showing acute infarcts (arrows) and thrombosed vessels (arrowhead) B A brain-blood vessel with a thrombus (arrowhead) isshown occluding the entire lumen Note the perivascular mononuclear inflammatory cells and towards the bottom of the image(arrow) necrotic cortical brain parenchyma (hematoxylin and eosin) C Higher magnification showing the organized thrombuswithin the blood vessel (hematoxylin and eosin) D Brainstem with acute microhemorrhage (hematoxylin and eosin)

TABLE 8 Summary of Microscopic Pathology Findings in the SpleenCOVID-19 SARS-CoV-1


White pulp depletion 81 18 (222) 23 23 (1000) SARS 45377 lt 0001Inflammation 78 15 (192) 23 0 (00) COVID 5195 0023Necrosis 81 5 (62) 23 6 (261) SARS 7511 0006Thrombi 78 0 (00) 23 0 (00) mdash mdash mdash




(TUNEL) DNA fragmentation assay38 No lymphocyticor neutrophilic infiltrate was reported Also no increasedparafollicular cells were seen

Musculoskeletal System Pathology FindingsSkeletal muscle involvement in SARS-CoV-1 was

reported by Leung et al39 with notable findings includingmyofiber necrosis macrophage infiltration and scant re-generative fibers COVID-19 studies did report onmusculoskeletal findings (Supplemental Digital ContentTable 8 httplinkslwwcomPASB89)

DISCUSSIONCOVID-19 caused by infection with SARS-CoV-2

can have variable clinical presentations ranging from amild asymptomatic respiratory illness to fulminantARDS multiorgan failure and death in a subset of pa-tients Given the pressing need to characterize the under-lying pathology associated with COVID-19 manypathologists spanning the globe bravely began performingautopsies and reporting their postmortem findings in thesedeceased patients This article to the best of our knowl-edge represents one of the largest systematic reviews andmeta-analysis of postmortem findings that have been re-ported associated with SARS-CoV-2 This review alsouniquely compares the pathology of COVID-19 to pre-viously published findings in SARS-CoV-1 associateddeaths There are several limitations of this study Giventhe rapid pace at which the literature is expanding on thistopic we may not have captured every article published onthe topic which includes newer articles Some of the ar-ticles included were preprint papers without peer reviewAlso as we relied on the data documented in these papersit was not always possible to extract all of the clinical

laboratory and pathology details desired nor the cause ofdeath or clinical correlation in all cases In addition theoverall number of published COVID-19 cases was higherthan that for SARS-CoV-1 which may have affectedcomparisons

Unlike SARS-CoV-1 where prior autopsy pub-lications arose largely from China with COVID-19 sucharticles spanned the globe which probably reflects thewidespread pandemic and high case fatality rate in manycountries Our analysis corroborates what has been re-ported in antemortem studies and large autopsy series79

which is that mortality due to COVID-19 occurs mostly ininfected patients who are elderly (median= 700 y) malesand those with an underlying comorbid disease such ashypertension diabetes andor obesity This is significantlydifferent from patients who succumbed to SARS-CoV-1as those patients were slightly younger and without suchreported comorbidities Of note entirely reporting co-morbid disease was more prevalent in COVID-19 articles

The most prominent pathologic manifestation ofCOVID-19 was the pulmonary findings which was likely themost common cause of symptoms and death in afflictedpatients1480ndash88 This review shows that the majority of patientswho died from COVID-19 had bilateral DAD similar toSARS-CoV-1 and MERS The histopathologic findings ofDAD from COVID-19 appear to be indistinguishable fromother causes of DAD89 The acuteexudative phase of DAD ischaracterized by inflammatory cell-mediated alveolar damagewith alveolar edema andor hemorrhage capillary congestionand hyaline membranes with or without microvascularthrombi The proliferativeorganizing phase of DAD showstype II pneumocyte hyperplasia reactive pneumocytes al-veolar wall thickening and myofibroblast proliferationwhereas the chronicfibrotic phase shows honeycomb lung

FIGURE 14 Hemophagocytic histiocytes (arrows) in COVID-19 autopsies from the bone marrow (A) subcapsular sinus of a lymphnode (B) and spleen (C) (hematoxylin and eosin)

TABLE 9 Summary of Microscopic Liver Pathology FindingsCOVID-19 SARS-CoV-1


Hepatocyte necrosis 85 16 (186) 15 11 (733) SARS 19221 lt 0001Portal inflammation 82 13 (159) 15 8 (533) SARS 10501 0001Thrombi 82 15 (183) 15 0 (00) COVID 3246 0072Lobular inflammation 82 3 (37) 15 0 (00) COVID 0566 0452




with collagenous fibrosis of alveolar spaces and an interstitiumwith thickening of the alveolar wall along with squamousmetaplasia of alveoli Of the typical phases of DAD (acuteexudative and proliferativeorganizing) these autopsy studiesrevealed that COVID-19 patients mostly showed lungs withan exudative phase compared with SARS-CoV-1 patientsindicating more early severe involvement of the lungs inCOVID-19 patients and earlier death Indeed this findingcorresponds to the average number of days from disease onsetto death which was up to 61 days for COVID-19 comparedwith 108 days for SARS-CoV-1 patients More recently fi-brosing DAD has been reported in patients with a longerduration of COVID-19 illness and hospitalization90 HoweverPolak et al91 showed that patients can also present with gt1pattern either simultaneously or consecutively The diagnosisof AFOP was mentioned in some articles as a prominentpattern of lung injury in COVID-19 patients AFOP can beseen in all forms of acute lung injury (DAD and acutepneumonia) Therefore in autopsy articles biased by core bi-opsies only the predominance of AFOP may be explained bythe limited sampling In COVID-19 pulmonary interstitialinflammation was surprisingly subtle which differs from otherviral infections where interstitial pneumonitis is typically aprominent feature92

Two notable postmortem findings reported in thelungs of COVID-19 patients typically not encountered inassociation with DAD include marked pulmonaryhemorrhage and scattered multinucleated giant cells Fatalhemorrhagic pneumonia has previously rarely been re-ported in association with other viral infections9394 Whilethe etiology of pulmonary hemorrhage is unclear studieshave suggested that in severe cases of COVID-associatedpneumonia there may be microvascular injury mediatedby activation of complement pathways41 Also the pos-sibility of therapy-induced hemorrhage especially due tohigh ventilation pressures in cases early in the pandemicmay have caused some of the hemorrhages if there is in-creased endothelial fragility in the setting of complementactivation The reason for multinucleated giant cell for-mation with COVID-19 is also unclear However similarsyncytial giant cells have been reported in the lungs withother viral infections and may represent a cellular reactionto a marked proinflammation cytokine-rich milieu95

The pathologic findings detected in other organs mayhelp explain some of the extrapulmonary manifestations ofCOVID-1996ndash99 Ancillary tests have detected coronaviruswithin most nonpulmonary organs including the heart kid-ney brain gastrointestinal tract liver spleen bone marrowlymph nodes and blood vessels and when present abovetypical levels of viremia raise the possibility of direct viralinfection of tissues12404448495355575965 SARS-CoV-2 wasalso isolated from the middle ear and mastoid in 2 autopsycases100 and was detected in the retina in 3 of 14 eyes fromdeceased COVID-19 patients101 The heart appears to be animportant target organ of SARS-CoV-2 because in some fatalinfections postmortem examination revealed individual my-ocyte necrosis myocarditis and increased interstitialmacrophages102 Clinically patients with COVID-19 havebeen reported to have electrocardiogram abnormalities in-

creased troponins and even more sinister cardiac-relatedevents such as cardiogenic shock103104 These findings raisesuspicion for targeting of the heart by SARS-CoV-2 Mostpathology studies have revealed nonspecific individual my-ocyte necrosis a finding common in critically ill patientsparticularly those with shock Whereas cases of COVID-19were identified with myocarditis and patchy mononuclearinfiltrates55 significant inflammation in the heart was not afinding reported with SARS-CoV-1 infection105 InterestinglyFarina et al106 suggested that the heart may be a source oflatent or persistent infection

The main histopathologic finding in the kidneysfrom patients with COVID-19 was acute tubular damageSuch acute tubular injury can lead to acute renal failure andcontribute to rapid clinical deterioration and mortality107

Neurological manifestations of COVID-19 (eg headacheanosmia or stroke among others) also appear to be commonWhile most postmortem central nervous system examinationshave revealed limited pathology a subset of studies showedpathology that included thrombi acute infarction punctatehemorrhage and hypoxiaischemia related changes108109

Such observations suggest the role of endothelial dysfunctionand thrombosis in neurological manifestations of COVID-19Thus far no postmortem reports of direct infectious compli-cations such as meningitisencephalitis have been identified atautopsy In addition to direct infection parainfectious com-plications in the setting of immune dysregulation remain apossibility as highlighted by the case report by Reichardet al110 that showed vascular and acute disseminatedencephalomyelitisndashlike pathology The presence of high levelsof coronavirus within the olfactory bulb supports viral entryinto the brain via the lamina cribrosa which may also explainthe clinical findings of smell dysfunction in COVID-19patients48 The hematological system was also affected inCOVID-19 Peripheral blood lymphopenia was a commonclinical feature of COVID-19 which corresponds to lymphoiddepletion observed in the lymph nodes and white pulp de-pletion of the spleen COVID-19 also causes gastrointestinalsymptoms such as diarrhea and abdominal pain Identificationof SARS-CoV-2 RNA within the gastrointestinal tract as wellas evidence of viral shedding in stool suggest direct gut in-fection with epithelial injury as a source of symptoms Ische-mic changes were also identified which could further explaingastrointestinal symptoms associated with COVID-19 al-though this is likely to occur later and in those with com-promised perfusion

Thrombosis was a common finding in almost all organsexamined in COVID-19 patients In the lungs thrombi weredescribed within all sized blood vessels ranging from capil-laries to the pulmonary artery and in one series they wereidentified bilaterally in over 25 of the pulmonaryparenchyma45 Using 3-dimensional imaging technology Liet al42 demonstrated that the extent of thrombi within smallvessels associated with COVID-19 is massive and greater thananticipated Microthrombi were also reported in the lungsfrom patients who died of SARS-CoV-1 On the basis of ul-trastructural studies alveolar-capillary microthrombi werereported to be 9 times more prevalent in patients with COV-ID-19 than in patients with influenza53 Pulmonary micro-




thrombi are known constituents of DAD encountered even innon-COVID conditions111112 Thus it is difficult to know ifpulmonary thrombi are indicative of DAD with or withoutadditional physiological derangements of critical illness in-cluding complement andor DIC In the brains of someCOVID-19 patients thrombi were associated with hemorrhageand acute infarction Small thrombi associated with COVID-19 have also been reported albeit infrequently in thedermis113 Some authors also report the presence of neu-trophilic plugs trapped in various organs114 Widespreadthrombotic disease clearly plays an important role in mor-bidity and more than likely accentuates mortality in these in-fected patients The etiology of thrombus formation in infectedindividuals is likely multifactorial with proposed mechanismsincluding microvascular dysfunction hyperimmune responseor dysregulation complement system dysregulation and al-tered coagulability including DIC all which may be furtherexacerbated by preexisting conditions23ndash25115ndash118 Anotherpathologic finding identified in several organs was vascularinjury Vascular injury specifically endotheliitis was recordedin both the COVID-19 and SARS-CoV-1 groups In COVID-19 cases this included lymphocytic inflammation cytoplasmicvacuolization endothelial detachment and microscopic de-struction of the blood vessel wall53 Vasculopathy is likelyascribed to a combination of direct viral injury inflammationandor complement-mediated vascular injury4158 Adrenalinsufficiency secondary to vasculopathy has been proposed asan important contributing factor to the cytokine storm ob-served in patients with severe COVID-19119

Additional systemic findings noted in the COVID-19group include features suggestive of direct viral cellular injury(eg candidate viral cytopathic effect and multinucleated giantcell formation) as well as indirect viral-induced pathology (eghemophagocytosis and megakaryocyte recruitment) Theviral-type cytopathic effect identified within scattered pneu-mocytes was much more common in COVID-19 than SARS-CoV-1 autopsies Interestingly this cytologic finding was notas well recognized in other organs This may be related to thefact that a greater number of ACE-2 receptors the cellularreceptor of SARS-CoV-2 have been reported on pneumocytesand endothelial cells53120 Some authors also report findingssimilar to smudge cells reminiscent of adenovirus-relatedpneumonitis121 A review article by Walsh and colleagueshighlights the timeline of viral load in the COVID-19 diseasecourse Viral load of SARS-CoV-2 from upper respiratorytract samples peaks around symptom onset or a few daysthereafter and then becomes undetectable about 2 weeks latercorresponding to the organizing phase of COVID-19 relatedpneumonia74122 Viral loads from sputum samples may con-tain higher levels of virus peak later and persist for longerThis timeline may be valuable for the managementmonitor-ing of COVID-19 pneumonia Prilutskiv et al123 report that intheir small series of 4 autopsies lymphophagocytosis appearsto be the predominant form of SARS-CoV-2 infection-asso-ciated hemophagocytosis observed The reason for having anincreased number of activated megakaryocytes in alveolarcapillaries is unclear but this phenomenon is believed to leadto increased platelet release with subsequent thromboticpathway activation43124125

In summary this systematic review and meta-anal-ysis reveal that overall autopsy findings are similar in bothCOVID-19 and SARS-CoV-1 groups Although the lungshad the most pronounced pathologic findings ex-trapulmonary disease was also identified but more varia-ble Similarities between COVID-19 and SARS-CoV-1were that the deceased were mostly adults who developedDAD However death due to COVID-19 tends to affectmore elderly men with preexisting comorbidities126 No-table pathologic findings associated with DAD especially inCOVID-19 include pulmonary hemorrhage multinucleatedgiant cells and viral cytopathic effect of pneumocytes Whilemost of the extrapulmonary histopathologic findings amongboth groups were not statistically significant with COVID-19there was more extensive thrombotic disease and multifocalendotheliitis Tissue inflammation was surprisingly under-whelming in most of the organs examined Further autopsiesare needed including correlation with clinical finings to betterelucidate the pathogenesis of SARS-CoV-2 infection and helpcontribute to the clinical management of infected patients

ACKNOWLEDGMENTSThe authors dedicate this publication to their late co-

author Dr Mary Fowkes whose scientific expertise andimages were foundational to this study Dr Fowkes whopassed away unexpectedly on November 15 2020 was arenowned Professor of Pathology and Director of Autopsyamp Neuropathology at Mount Sinai HospitalIcahn Schoolof Medicine New York NY She worked tirelessly duringthe COVID-19 pandemic performing COVID-19 autopsiesand revolutionizing the COVID autopsy service at MountSinai Hospital Her autopsy findings helped to improveclinical care for COVID-19 patients and she received nationalrecognition for this heroic work during an appearance on CNNAnderson Cooper-3600 She was a member of the Board ofGovernors of CAP and the author of multiple publications in thearea of neuropathology The authors are forever personally andprofessionally indebted to Dr Fowkes for her expertise andfriendship They also thank Colleen Vrbin from Analytical In-sights for help with statistical analysis

REFERENCES1 Hwang DM Chamberlain DW Poutanen S et al Pulmonary

pathology of severe acute respiratory syndrome in Toronto ModPathol 2005181ndash10

2 Ng DL AlHosani F Keating MK et al Clinicopathologicimmunohistochemical and ultrastructural findings of a fatal caseof middle East respiratory syndrome coronavirus infection in theUnited Arab Emirates April 2014 Am J Pathol 2016186652ndash658

3 Huang C Wang Y Li X et al Clinical features of patients infected with2019 novel coronavirus in Wuhan China Lancet 2020395497ndash506

4 Worldometers COVID-19 coronavirus pandemic 2020 Availableat wwwworldometersinfocoronavirus Accessed May 17 2020

5 Worldometers COVID-19 coronavirus pandemic 2020 Availableat wwwworldometersinfocoronavirus Accessed July 17 2020

6 Rabaan AA Al-Ahmed SH Haque S et al SARS-CoV-2 SARS-CoVand MERS-CoV a comparative overview Infez Med 20202174ndash184

7 Gattinoni L Chiumello D Rossi S COVID-19 pneumonia ARDSor not Crit Care 202024154

8 Gattinoni L Chiumello D Caironi P et al COVID-19 pneumoniadifferent respiratory treatments for different phenotypes IntensiveCare Med 2020461099ndash1102




9 Danzi GB Loffi M Galeazzi G et al Acute pulmonary embolismand COVID-19 pneumonia a random association Eur Heart J2020411858

10 Liu Q Wang RS Qu GQ et al Gross examination report of aCOVID-19 death autopsy Fa Yi Xue Za Zhi 20203621ndash23

11 Xiaohong Y Tingyuan L Zhicheng H et al New coronaviruspneumonia (COVID-19) 3 cases of multiple-site puncture histopatho-logical pathology of human remains Chin J Pathol 202049291ndash293

12 Tian S Xiong Y Liu H et al Pathological study of the 2019novelcoronavirus disease (COVID-19) through postmortem corebiopsies Mod Pathol 2020331007ndash1014

13 Konopka KE Wilson A Myers J Postmortem lung findings in anasthmatic with coronavirus disease 2019 (COVID-19) Chest2020158E99ndashE101

14 Fox SE Akmatbekov A Harbert JL et al Pulmonary and cardiacpathology in COVID-19 the first autopsy series from New OrleansLancet Respir Med 20208681ndash686

15 Salerno M Sessa F Piscopo A et al No autopsies on COVID-19deaths a missed opportunity and the lockdown of science J ClinMed 202091472

16 Pomara C Volti GL Cappello F COVID-19 deaths are we sure it ispneumonia Please autopsy autopsy autopsy J ClinMed 202091259

17 Barth RF Colombus RH Xinyang X et al A call to action theneed for autopsies to determine the full extent of organ involvementassociated with COVID-19 Chest 202015843ndash44

18 Zhou B Zhang W Feng R et al The pathological autopsy ofcoronavirus disease 2019 (COVID-2019) in China a review PathogDis 202078ftaa026

19 Occupational Safety and Health Administration (OSHA) Availableat wwwoshagov Accessed May 17 2020

20 Lacy JM Brooks EG Akers J et al COVID-19 postmortemdiagnostic and biosafety considerations Am J Forensic Med Pathol202041143ndash151

21 Williamson AK Creation and benefits of the ldquoCOVID AutopsyListserverdquo Arch Pathol Lab Med 20201441160ndash1161

22 Lillicrap D Disseminated intravascular coagulation in patients with2019-nCoV pneumonia J Thromb Haemost 202018786ndash787

23 Henry BM Vikse J Benoit S et al Hyperinflammation derange-ment of renin-angiotensin-aldosterone system in COVID-19 anovel hypothesis for clinically suspected hypercoagulopathy andmicrovascular thrombosis Clin Chim Acta 2020507167ndash173

24 Yang X Yu Y Xu J et al Clinical course and outcomes ofcritically ill patients with SARS-CoV-2 pneumonia in WuhanChina a single-centered retrospective observational study LancetRespir Med 20205475ndash481

25 Zhou F Yu T Du R et al Clinical course and risk factors formortality of adult in patients with COVID-19 in Wuhan China aretrospective cohort study Lancet 20203951054ndash1062

26 Franks TJ Chong PY Chui P et al Lung pathology of severeacute respiratory syndrome (SARS) a study of 8 autopsy cases fromSingapore Hum Pathol 200334743ndash748

27 Chong PY Chui P Lui AE et al Analysis of deaths during thesevere acute respiratory syndrome (SARS) epidemic in SingaporeArch Pathol Lab Med 2004128195ndash204

28 Nicholls JM Poon LLM Lee KC et al Lung pathology of fatalsevere acute respiratory syndrome Lancet 20033611773ndash1778

29 Gu J Gong E Zhang B et al Multiple organ infection and thepathogenesis of SARS J Exp Med 2005202415ndash424

30 Lai RQ Feng XD Wang ZC et al Pathological and ultra-microstructural changes of tissues in a patient with severe acuterespiratory syndrome Chin J Pathol 200332205ndash208

31 Chen J Xie YQ Zhang HT et al Lung pathology of severe acuterespiratory syndrome Chin J Pathol 200325360ndash362

32 Lang ZW Zhang LJ Zhang SJ et al A clinicopathological studyon 3 cases of severe acute respiratory syndrome Chin J Pathol200332201ndash204

33 Hsiao CH Wu MZ Hsies SW et al Clinicopathology of severeacute respiratory syndrome an autopsy case report J Formos MedAssoc 2004103787ndash792

34 Ding Y He L Zhang Q et al Organ distribution of severe acuterespiratory syndrome (SARS) associated coronavirus (SARS-CoV)

in SARS patients implications for pathogenesis and virus trans-mission pathways J Pathol 2004622ndash630

35 Tang JW To KF Lo AWI et al Quantitative temporal-spatialdistribution of severe acute respiratory syndrome-associated coro-navirus (SARS-CoV) in post-mortem tissues J Med Virol2007791245ndash1253

36 Zhan J Deng R Tang J et al The spleen as a target in severe acuterespiratory syndrome FASEB J 2006202321ndash2328

37 Shi X Gong E Gao D et al Severe acute respiratory syndromeassociated coronavirus is detected in intestinal tissues of fatal casesAm J Gastroenterol 2005100169ndash176

38 Wei L Sun S Zhu CH et al Pathology of the thyroid in severeacute respiratory syndrome Hum Pathol 20073895ndash102

39 Leung TW Wong KS Hui AC et al Myopathic changesassociated with severe acute respiratory syndrome Arch Neurol2005621113ndash1117

40 Bradley BT Maioli H Johnson R et al Histopathology andultrastructural findings of fatal COVID-19 infections in Washing-ton State a case series Lancet 2020396320ndash332

41 Magro C Mulvey JJ Berlin D et al Complement associatedmicrovascular injury and thrombosis in the pathogenesis of severeCOVID-19infection a report of five cases Transl Res 20202201ndash13

42 Li G Fox SE Summa B et al Multiscale 3-dimensional pathologyfindings of COVID-19 diseased lung using high-resolution clearedtissue microscopy bioRxiv preprint 2020

43 Barton LM Duval EJ Stroberg E et al COVID-19 autopsiesOklahoma USA Am J Clin Pathol 2020153725ndash733

44 Varga Z Flammer AJ Steiger P et al Endothelial cell infectionand endotheliitis in COVID-19 Lancet 20203951417ndash1418

45 Carsana L Sonzogni A Nasr A et al Pulmonary post-mortemfindings in a series of COVID-19 cases from Northern Italy a two-centre descriptive study Lancet Infect Dis 2020201135ndash1140

46 Wichmann D Sperhake JP Lutgehetmann M et al Autopsyfindings and venous thromboembolism in patients with COVID-19Ann Intern Med 2020173268ndash277

47 Karami P Naghavi M Feyzi A et al Mortality of a pregnantpatient diagnosed with COVID-19 a case report with clinicalradiological and histopathological findings Travel Med Infect Dis2020 [Epub ahead of print]

48 Menter T Haslbauer JD Neinhold R et al Post-mortem examinationsof COVID-19 patients reveals diffuse alveolar damage with severecapillary congestion and variegated findings in lungs and other organssuggesting vascular dysfunction Histopathology 202077198ndash209

49 Farkash EA Wilson AM Jentzen JM Ultrastructral evidence fordirect renal infection with SARS-CoV-2 J Am Soc Nephrol2020311683ndash1687

50 Craver R Huber S Sandomirsky M et al Fatal eosinophilicmyocarditis in a healthy 17-year-old male with severe acuterespiratory syndrome coronavirus 2 (SARS-CoV-2c) Fetal PediatrPathol 202039263ndash268

51 Lei Y Sun S Xu CH et al COVID-19 in a Hispanic womanautopsy report with clinical pathological correlation Arch PatholLab Med 20201441041ndash1047

52 Lax SF Skok K Zechner P et al Pulmonary arterial thrombosis inCOVID-19 with fatal outcome results from a prospective single-centerclinicopathologic case series Ann Intern Med 2020173350ndash361

53 Ackermann M Verleden SE Kuehnel M et al Pulmonary vascularendothelialitis thrombosis and angiogenesis in COVID-19 N EnglJ Med 2020383120ndash128

54 Adachi T Chong JM Nakajima M et al Clinicopathologic andimmunohistochemical findings from autopsy of patient withCOVID-19 Japan Emerg Infect Dis 2020262157ndash2161

55 Bryce C Grimes Z Pujadas E et al Pathophysiology of SARS-CoV-2 targeting of endothelial cells renders a complex disease withthrombotic microangiopathy and aberrant immune response TheMount Sinai COVID-19 autopsy experience Available at httpsdoiorg1011012020051820099960 Accessed May 17 2020

56 Copin MC Parmetier E Dubuurcq T et al Time to considerhistologic pattern of lung injury to treat critically ill patients withCOVID-19 infection Intensive Care Med 2020461124ndash1126




57 Su H Yang M Wan C et al Renal histopathological analysis of 26postmortem findings of patients with COVID-19 in China KidneyInt 2020219219ndash227

58 Diao B Feng Z Wang C et al Human kidney is a target for novelsevere acute respiratory syndrome coronavirus 2 (SARS-CoV-2)infection Available at httpsdoiorg1011012020030420031120Accessed May 17 2020

59 Xia X Xiaona C Huaxiang P et al Pathological changes of the spleenin ten patients with coronavirus disease 2019(COVID-19) by postmor-tem needle autopsy Zhonghua Bing Li Xue Za Zhi 202049576ndash582

60 DerSimonian R Laird N Meta-analysis in clinical trials revisitedContemp Clin Trials 201545139ndash145

61 Freeman MF Tukey JW Transformations related to the angularand the square root Ann Math Statist 195021607ndash611

62 Egger M Smith GD Schneider M et al Bias in meta-analysisdetected by a simple graphical test BMJ 1997315629ndash634

63 Hanley B Lucas SB Youd E et al Autopsy in suspected COVID-19 cases J Clin Pathol 202073239ndash242

64 Centers for Disease Control and Prevention Collection andsubmission of postmortem specimens from deceased personswith known or suspected COVID-19 interim guidance March2020 Available at wwwcdcgovcoronavirus Accessed May 172020

65 Public Health England COVID-19 guidance for sampling and fordiagnostic laboratories secondary COVID-19 guidance for sam-pling and for diagnostic laboratories 2020 Available at www govukgovernmentpublicationswuhan novel-coronavirus-guidance-for-clinical-diagnostic-laboratories Accessed May 17 2020

66 Fineschi V Aprile A Aquila I et al Management of the corpsewith suspectprobable or confirmed COVID-19 respiratory infectionndash Italianinterim recommendations for personnel potentially exposedto materialfrom corpses including body fluids in morgue structuresandduring autopsy practice Pathologica 202011264ndash67

67 The Royal College of Pathologists Autopsy practice relatingtopossible cases of COVID-19 (2019-nCov novel coronavirus fromChina 20192020) 2020 Available at wwwrcpathorg AccessedMay 17 2020

68 Dell-Aquilla M Cattani P Fantoni M et al Postmortem swabs inthe Sars-CoV-2 pandemic report on 12 complete clinical autopsycases Arch Pathol Lab Med 20201441298ndash1302

69 Youd E Moore L COVID-19 autopsy in people who died incommunity settings the first series J Clin Pathol 202073840ndash844

70 Hasmi AH Khoo LS Koo ZP et al The craniotomy box aninnovative method of containing hazardous aerosols generatedduring skull saw use in autopsy on a COVID-19 body Forensic SciMed Pathol 202016477ndash480

71 Hofman P Puchois P Brest P et al Possible consequences of theCOVID-19 pandemic on the use of biospecimens from cancerbiobanks for research in academia and bioindustry Nat Med202026890ndash898

72 Calabrese F Pezzuto F Fortarezza F et al Pulmonary pathologyand COVID-19 lessons from autopsy The experience of EuropeanPulmonary Pathologists Virchows Arch 2020477359ndash372

73 Borczuk AC Salvatore SP Seshan SV et al COVID-19 pulmonaryparthology a multi-institutional autopsy cohort from Italy andNew York City Mod Pathol 2020332156ndash2168

74 Schaefer IM Padera RF Solomon IH In situ detection of SARS-CoV-2 in lungs and airways of patients with COVID-19 ModPathol 2020332104ndash2114

75 Bishop JA Ma XJ Wang H et al Detection of transcriptionallyactive high-risk HPV in patients with head and neck squamous cellcarcinoma as visualized by a novel E6E7 mRNA in situ hybrid-ization method Am J Surg Pathol 201231874ndash1882

76 Alsaad KO Hajeer AH Al Balwi M et al Histopathology ofMiddle East respiratory syndrome coronavirus (MERS-CoV)infectionmdashclinicopathological and ultrastructural study Histopa-thology 201872516ndash524

77 Yang M Chen S Huang B et al Pathological findings in the testesof COVID-19 patients clinical implications Eur Urol Focus202061124ndash1129

78 Elsoukkary SS Mostyka M Dillard A et al Autopsy findings in 32patients with COVID-19 a single-institution experience Pathobi-ology 2020 [Epub ahead of print]

79 Edler C Schroder AS Aepfelbacher M et al Dying with SARS-CoV-2 infection-an autopsy study of the first consecutive 80 cases inHamburg Germany Int J Legal Med 20201341275ndash1284

80 Barth RF Buja LM Parwani AV The spectrum of pathologicalfindings in coronavirus disease (COVID-19) and the pathogenesis ofSARS-CoV-2 Diagn Pathol 20201585

81 Bosmuller H Traxler S Bitzer M et al The evolution ofpulmonary pathology in fatal COVID-19 disease an autopsy studywith clinical correlation Virchows Arch 2020477349ndash357

82 Hariri LP North CM Shih AR et al Lung histopathology inCOVID-19 as compared to SARS and H1N1 influenza a system-atic review Chest 202015973ndash84

83 Maiese A Manetti AC La Russa R et al Autopsy findings inCOVID-19 related deaths a literature review Forensic Sci MedPathol 2020 [Epub ahead of print]

84 De Michele S Sun Y Yilmaz MM et al Forty postmortem examin-ations in COVID-19 patients Am J Clin Pathol 2020154748ndash760

85 Kommoss FKF Schwab C Taverner L et al The pathology ofsevere COVID-19 related lung damage Dtsch Arztebl Int2020117500ndash506

86 Eketunde AO Mellacheruvu SP Oreoluwa P et al A review ofpostmortem findings in patients with COVID-19 Cureus 202012e9438

87 Grosse C Gross A Salzer HJF et al Analysis of cardiopulmonaryfidnings in COVID-19 fatalities high incidence of pulmonary arterythrombi and acure suppurative bronchopneumonia CardiovascPathol 202049107263

88 AL Nemer A Histopathologic and autopsy findings in patientsdiagnosed with coronavirus disease 2019 (COVID-19) what weknow so far based on correlation with clinical morphologic andpathobiological aspects Adv Anat Pathol 202027363ndash370

89 Konopka KE Nguyen T Jentzen JM et al Diffuse alveolardamage (DAD) from coronavirus disease 2019 infection ismorphologically indistinguishable from other causes of DADHistopathology 202077570ndash578

90 Li Y Wu J Wang S et al Progression to fibrosing diffuse alveolardamage in a series of 30 minimally invasive autopsies with COVID-19 pneumonia in Wuhan China Histopathology 2020 Doi101111his14249

91 Polak SB Van Gool IC Cohen D et al A systematic review ofpathological findings in COVID-19 a pathophysiological timelineand possible mechanisms of disease progression Mod Pathol2020332128ndash2138

92 Radoycich GE Zuppan CW Weeks DA et al Patterns of measlespneumonitis Pediatr Pathol 199212773ndash786

93 Donoso AF Leoacuten JA Camacho JF et al Fatal hemorrhagicpneumonia caused by human metapneumovirus in an immuno-competent child Pediatr Int 200850589ndash591

94 Gilbert CR Vipul K Baram M Novel H1N1 influenza A viralinfection complicated by alveolar hemorrhage Respir Care 201055623ndash625

95 Villenave R Thavagnanam S Sarlang S et al In vitro modeling ofrespiratory syncytial virus infection of pediatric bronchial epithe-lium the primary target of infection in vivo Proc Natl Acad SciUSA 20121095040ndash5045

96 Gupta A Madhavan MV Sehgal K et al Extrapulmonarymanifestations of COVID-19 Nat Med 2020261017ndash1032

97 Deshmukh V Motwani R Kumar A et al Histopathologicalobservations in COVID-19 a systematic review J Clin Pathol2020 Doi 101136jclinpath-2020-206995

98 Mohanty SK Satapathy A Naidu MM et al Severe acuterespiratory syndrome coronavirus-2 (SARS-CoV-2) and coronavi-rus disease 19 (COVID-19)mdashanatomic pathology perspective oncurrent knowledge Diagn Pathol 202015103

99 Remmelink M De Mendonca R DrsquoHaene N et al Unspecificpost-mortem findings despite multiorgan viral spread in COVID-19patients Crit Care 202024495




100 Frazier KM Hooper JE Mostafa HH et al SARS-CoV-2 virusisolated from the mastoid and middle ear implications for COVID-19 precautions during ear surgery JAMA Otolaryngol Head NeckSurg 2020146964ndash966

101 Casgrande M Fitzek A Puschel K et al Detection of SARS-CoV-2in human retinal biopsies of deceased COVID-19 patients OculImmunol Inflamm 202028721ndash725

102 Basso C Leone O Rizzo S et al Pathological features of COVID-19-associated myocardial injury a mutlicentre cardiovascularpathology study Eur Heart J 2020413827ndash3835

103 Buja LM Wolf DA Zhao B et al The emerging spectrum ofcardiopulmonary pathology of the coronavirus disease 2019(COVID-19) report of 3 autopsies from Houston Texas andreview of autopsy findings from other United States citiesCardiovasc Pathol 202048107233

104 Tavazzi G Pellegrini C Maurelli M et al Myocardial localizationof coronavirus in COVID-19 cardiogenic shock Eur J Heart Fail202022911ndash915

105 Fox SE Li G Akmatbekoy A et al Unexpected features of cardiacpathology in COVID-19 infection Circulation 20201421123ndash1125

106 Farina A Uccello G Spreafico M et al SARS CoV-2 detection inthe pericardial fluid of a patient with cardiac tamponade Eur JIntern Med 202026100ndash101

107 Santoriello D Khairallah P Bomback AS et al Postmortem kidneyfindings in patients with COVID-19 J Am Soc Nephrol 2020312158ndash2167

108 Al-Siraj S Troakes C Hanley B et al The spectrum ofneuropathology in COVID-19 Neuropathol Appl Neurobiol 2020Doi 101111nan12667

109 Kantonen J Mahzabin S Mayranpaa MI et al Nueropathologicfeatures of four autopsied COVID-19 patients Brain Pathol 2020301012ndash1016

110 Reichard RR Kashani KB Boire NA et al Neuropathology ofCOVID-19 a spectrum of vascular and acute disseminated encephalo-myelitis (ADEM)-like pathology Acta Neuropathol 20201401ndash6

111 Tomashefski JF Davies P Boggis C et al The pulmonary vascularlesions of the adult respiratory distress syndrome Am J Pathol1983112112ndash126

112 Harms PW Schimdt LA Smith LB et al Autopsy findings in eightpatients with fatal H1N1 influenza Am J Clin Pathol 201013427ndash35

113 Daurte-Neto AN Almeida-Monteiro RA Ferraz-da-Silva LFet al Pulmonary and systemic involvement of COVID-19 assessedby ultrasound-guided minimally invasive autopsy Histopathology202077186ndash197

114 Schurink B Roos E Radonic T et al Viral presence andimmunopathology in patients with lethal COVID-19 a prospectiveautopsy cohort study Lancet Microbe 20201E290ndashE299

115 Rosenthal FR Causes of venous thrombosis Thromb J 201614(suppl 1)24

116 Foley JH Examining coagulation-complement crosstalk comple-ment activation and thrombosis Thromb Res 2016141(suppl 2)S50ndashS54

117 Rizzo P Sega FVD Fortini F et al COVID-19 in the heart and thelungscould we ldquoNotchrdquo the inflammatory storm Basic ResCardiol 202011531

118 Lo MW Kemper C Woodruff TM COVID-19 complementcoagulation and collateral damage J Immunol 20202051488ndash1495

119 Iuga AC Marboe CC Yilmaz MM et al Adrenal vascular changesin COVID-19 autopsies Arch Pathol Lab Med 20201441159ndash1160

120 Yuki K Jufiogi M Koutsogiannaki S COVID-19 pathophysio-logy a review Clin Immunol 2020215108427

121 Conde PN Monraval PA Medina CM et al Autopsy findingsfrom the first known death from severe acute respiratory syndromeSARS-CoV-2 in Spain Rev Esp Pathol 202053188ndash192

122 Walsh KA Jordan K Clyne B et al SARS-CoV-2 detection viralload and infectivity over the course of an infection J Infect202081357ndash371

123 Prilutskiv A Kritselis M Shevotsov A et al SARS-CoV-2infection-associated hemophagocytic lymphohistiocytosis Am JClin Pathol 2020154466ndash474

124 Falasca L Nardacci R Colombo D et al Postmortem findings inItalian patients with COVID-19 a descriptive full autopsy study of caseswith and without co-morbidies J Infect Dis 20202221807ndash1815

125 Rapkiewicz AV Mai X Carsons SE et al Megakryocytes andplatelet-fibrin thrombi characterize multi-organ thrombosis atsutopsy in COVID-19 a case series EClinicalMedicine 202024100434

126 Khalil M Allen TC Krdowska A Acute respiratory distresssyndrome a review article AJSP Rev Rep 2017223ndash7




Studies have shown that SARS-CoV-2 is an RNAvirus that belongs to the same family of beta coronavirusesas SARS-CoV-1 and MERS-CoV Minor differences inthe spike protein of SARS-CoV-2 are associated with itsincreased transmission rate6 Both SARS-CoV-1 andSARS-CoV-2 use the angiotensin-converting enzyme 2(ACE-2) receptor for binding to host cells It is speculatedthat the pathogenesis of COVID-19 is similar to that ofSARS Although acute respiratory distress syndrome(ARDS) and its histologic hallmark DAD is consideredto be the most common cause of death in COVID-19 thepathologic basis of why ARDS occurs only in a subset ofpatients and why only a subset of these afflicted in-dividuals die is unclear78 Furthermore the clinical courseof severe and fatally ill patients with COVID-19 has beencomplicated by cardiac injury shock and thrombosis in-cluding pulmonary embolism and stroke9

The literature regarding COVID-19 is rapidlyemerging with articles progressively unraveling thepathogenesis and underlying pathology of this infectiousdisease based largely on either gross examination onlylimited autopsy case reports small postmortem series andpartial autopsy findings10ndash14 Despite the active appeal toperform autopsies on COVID-19 patients the currentoverall autopsy rate appears to be low15ndash18 The paucity ofautopsy data is not surprising because initially there wererecommendations to suspend postmortems on patientswith suspectedconfirmed COVID-19 infection19 Alsoperforming postmortem examinations on COVID-19 pa-tients is risky and requires an autopsy facility with ap-propriate biosafety accessories20 Fortunately there hasbeen a recent drive to boost postmortem services duringthis pandemic21 This is important because many criticalquestions about COVID-19 remain to be answered Forexample apart from ARDS what role do other processessuch as direct nonpulmonary tissue viral infection im-mune dysregulation primary thrombosis disseminatedintravascular coagulation (DIC) and secondary infectionsplay in fatal COVID-19 disease22ndash25 Clearly there is anurgent need to characterize the underlying pathology ofCOVID-19 to better understand the pathogenesis of thisdisease and more effectively treat infected patients

The aim of this review article was to systematicallysummarize the available emerging literature regardingCOVID-19 autopsies The intent was to additionallyconduct a meta-analysis to compare postmortem findingsin decedents of COVID-19 to previously published find-ings in SARS-CoV-1 associated deaths

MATERIALS AND METHODSA review of the English and non-English literature in

online archival databases up to June 2020 was performedThere were 15 articles about autopsies on patients withconfirmed SARS-CoV-1126ndash39 and 24 articles detailingautopsies on patients with confirmed COVID-1911ndash1440ndash59

that were included (Supplemental Digital Content Table 1httplinkslwwcomPASB82) Two of the SARS-CoV-1articles referred to the same group of 8 patients2627 but

because they provided different clinical and pathologicdetails in each paper both were included without dupli-cating the patient count A total of 332 patients wereidentified including 91 with SARS-CoV-1 infection and241 with COVID-19 All of the papers concerning SARS-CoV-1 were previously published whereas only 15(625) of the papers about COVID-19 autopsies wereaccepted in the peer-reviewed literature at the time ofsubmitting this review with the remaining preprint articlesanalyzed available on medRxiv or bioRxiv One COVID-19 paper of a pregnant woman with COVID-19 has sincebeen withdrawn by the authors47

Specific patient age was provided for 131 patients(60 with SARS-CoV-1 71 with COVID-19) allowing foran age comparison for both patient groups The normalityof the distributions of patient age was examined using theShapiro-Wilk normality test As the data were not nor-mally distributed the Mann-Whitney test was used tocompare SARS-CoV-1 and COVID-19 patient ages Onearticle provided age ranges for each of the 11 patients intheir study52 and these were added to a categorical sum-mary of age with the other 131 patients The Pearson χ2test was used to determine if the proportion of cases withSARS-CoV-1 or COVID-19 was different among catego-rical data variables (sex age category raceethnicitydocumented comorbidities and microscopic pathologicdetails including inflammation thromboembolic findings andnecrosis) Pooled prevalence and their 95 confidence inter-vals of histopathologic findings was calculated using theDerSimonian-Laird random-effects model after double-arcsine transformation Double-arcsine transformation is theappropriate choice when the sample size is small andor ex-treme proportions need to be handled6061 A formal Egger testdetected no publication bias62 Statistical significance was as-sumed at P-value lt005 Analyses were performed using IBMSPSS Statistics 22 and software R 362 (R Foundation forStatistical Computing Vienna Austria) with R Studio125033 environment (R Studio Inc Boston MA)

The selected images included in this manuscript wereacquired from a total of 157 autopsies performed at theaforementioned institutions as well as from 26 personalconsultation cases of autopsies performed at other in-stitutions on patients with confirmed SARS-CoV-2

Autopsy ProcedureSince the outbreak of COVID-19 several articles have

emphasized the need to conduct autopsies using standardizedprotocols and recommended using at least a Biosafety Level-2 autopsy facility2063ndash65 Guidelines were also proposed forforensic pathology66 Although performing a complete bodyautopsy is ideal alternative options utilized to minimize therisk of SARS-CoV-2 transmission to autopsy staff have in-cluded performing limited postmortems with in situ samplingor minimally invasive puncturecore biopsy autopsies111267

of only major target organs with or without opening theentire body Autopsy procedure recommendations have in-cluded utilizing adequately trained staff minimizingthe number of people in the autopsy suite and allowing onlyone person to handle knives The autopsy facility should











n ()


SARS-CoV-1(N= 91)






n ()

Age(y)

COVID-19(N= 82)




























































































































































































































































































n ()


SARS-CoV-1(N= 91)






n ()

Age(y)

COVID-19(N= 82)





















































































































































































































































































postmortem findings associated with sars-cov-2

Documents