Study design
In this post-mortem study, we included the first 17 adult patients (> 18 years) who died in our hospital (either in a COVID-19 unit or an intensive care unit) from March 13, 2020, with confirmed SARS-CoV-2 infection (i.e., positive RT-PCR assay on nasopharyngeal swab and/or broncho-alveolar lavage specimen). Exclusion criteria were lack of family consent and a delay of more than 5 days after death before post-mortem examination. The study protocol was approved by the local ethics committee (P2020/218).
Data collection
We collected demographics, comorbidities, relevant clinical data, including duration between symptom onset or hospitalization and death, the results of chest computed tomography scan, and, if available, microbiological tests and medical treatments (e.g., hydroxychloroquine, antivirals or antibiotics, and use of organ support). Acute respiratory distress syndrome (ARDS) and acute kidney injury (AKI) were defined according to standard definitions [8, 9].
Post-mortem procedure
The Belgian Public Health Institute (Sciensano) guidelines were integrated into our post-mortem procedure [10]. The cadavers were kept in the refrigerator at 4 °C and autopsies were performed 72 to 96 h after death to ensure the safety of the autopsy team. Personal protective equipment consisted of two superposed disposable latex gloves, plastic sleeves, FFP3 mask, scrub hat, clear face visor, surgical gown plus plastic apron, and rubber boots. In the post-mortem room, dirty and clean circulations were used in the airlocks to allow decontamination. All analyses were performed at normal pressure.
Using standard surgical pathology processing, complete sets of tissue samples were collected for diagnosis and biobanking. The material was biobanked by Biobanque Hôpital Erasme-ULB (BE_BERA1), CUB Hôpital Erasme; BBMRI-ERIC. The banked material consists of 6 samples per organ, including the trachea, thyroid, lymph nodes, heart, spleen, bone marrow, kidney, bladder, liver, stomach, colon, and brain. For the lungs, we collected six samples per lobe (i.e., a total of 30 samples), except for two patients who had undergone lobectomy for cancer and from whom only 18 samples were taken. For safety reasons, complete brain removal was not allowed, but, with the help of a neurosurgeon, in 11 cases, we used a new, safe procedure with drills and protective devices to avoid air dispersion, to obtain between 12 and 51 samples from different brain regions, as detailed in the Additional file 1 (Additional Material). Formalin-fixed paraffin-embedded (FFPE) tissues underwent standard processing to provide hematoxylin and eosin (H&E)-stained sections. Special stains and immunohistochemistry (IHC) were used for lung (Masson’s trichrome, periodic acid-Schiff [PAS], Gomori-Grocott, anti-CMV IHC, anti-HSV IHC, anti-Pneumocystis J IHC) and kidney (PAS, Masson’s trichrome, Jones methenamine silver) samples.
Morphological analysis
Morphological analysis was performed on H&E stained glass slides using the SecundOs digital platform (TribVn Health Care, Chatillon, France) for digital diagnosis, after the acquisition of whole slide digital scans (× 40 magnification) using a Nanozoomer 2.0 HT slide scanner (Hamamatsu, Hamamatsu City, Japan).
SARS-CoV-2 detection by immunohistochemistry
Since no antibody against SARS-CoV-2 has been validated for IHC on FFPE tissues, we selected an anti-SARS-nucleocapsid protein antibody. Standard IHC was applied as previously described to 4-μm-thick post-mortem lung sections (one sample for each lung lobe per patient) to display SARS-nucleocapsid protein (Invitrogen, PA1-41098, dilution 1:50) on Dako Omnis (Agilent Technologies, Santa Clara, CA, USA) using the Envision Flex detection system according to the manufacturer’s protocol [11]. The sections were counterstained with hematoxylin. Negative tissue controls were obtained from patients who had an autopsy before the COVID-19 pandemic. Semi-quantitative IHC evaluation was performed by two senior pathologists (ND, MR) as follows: negative (−); between one and five positive cells per whole slide (scattered cells, +); more than five cells per whole slide but no foci (isolated cells, ++); and with foci (more than 10 cells in one × 20 field, +++).
SARS-CoV-2 detection by rRT-PCR
Total nucleic acid was extracted from FFPE tissues using the Maxwell RSC DNA FFPE Kit (reference: AS1450. Promega Corporation, Madison, WI, USA) and the Promega Maxwell extractor, following the protocol described by the manufacturer. One-step RT-PCR assays specific for the amplification of SARS-CoV-2 E envelope protein gene were adapted from a published protocol [12]. Briefly, 4 μL of RNA (100 ng) was amplified in 20 μL reaction mixture containing 5 μL of TaqMan Fast Virus 1-step master mix (Life Technologies), 0.4 μM of each forward (ACAGGTACGTTAATAGTTAATAGCGT) and reverse (ATATTGCAGCAGTACGCACACA) primers and 0.2 μM of probe (FAM-ACACTAGCCATCCTTACTGCGCTTCG-BBQ). The amplification condition was 50 °C for 10 min for reverse transcription, followed by 95 °C for 20 s and then 45 cycles of 95 °C for 3 s and 58 °C for 30 s. A clinical sample highly positive for SARS-CoV-2 was diluted 1:1000 and used as a positive control in each analysis. A clinical sample obtained from a patient who was autopsied before the COVID-19 pandemic was used as a negative control. The quality of the RNA from the samples showing negative results was assessed by amplification of the human MET RNA according to a validated ISO:15189 accredited method used as a routine diagnostic method in our laboratory.
Statistical analysis
Data are reported as counts (percentage) or medians [interquartile ranges (IQRs)]. All data were analyzed using GraphPad Prism Version 8.4.2 (GraphPad Software, San Diego, CA, USA).