Of the examined 80 deceased with SARS-CoV-2 infection, all but the first autopsy, which had been requested by relatives, were ordered by the public health department. The autopsies were performed on average 4 days after death.
The deceased were aged between 52 and 96 years (average 79.2 years, median 82.4 years). Thirty-four of the deceased were female (42%) and 46 male (58%). In twelve cases, the place of death was the patient’s own home (15%); in 51 cases, the hospital (64%); in thirteen cases, a nursing or retirement home (16%); and in one case, a hotel. In three cases, no information was available on the place of death. Of the patients who died in hospitals, 17 died in intensive care units (ICU) with invasive ventilation, 31 in a normal ward, and one in the emergency room. The exact place of death could not be determined with regard to two patients.
The average body mass index (BMI) was 25.9 kg/m2. Overall, 38% of the deceased were overweight or obese (overweight 13 cases, obesity grade 1 six cases, grade 2 five cases, grade 3 six cases).
All deceased, except for two women, in whom no significant pre-existing conditions were seen autoptically (cases 7 and 50), had relevant previous illnesses. The vast majority had diseases of the cardiovascular system (85%), followed by lung diseases (55%), kidney diseases (34%), and central nervous system (CNS) diseases (35%). Diabetes mellitus was known in 21% of the deceased, and carcinomas/haematological diseases in 16%. Table 2 gives a systematic overview of all the cases.
Table 2 Autopsy cases of SARS-CoV-2 positive deaths In six of the deceased, SARS-CoV-2 infection was diagnosed postmortem. In the other 74 cases, SARS-CoV-2 infection had already been known antemortem. The time of the first positive PCR test is known in 49 cases. The average survival time after the first positive test until death was 6 days. The longest documented survival time was 32 days in the case of an 89-year-old man who died in a normal hospital ward.
In most cases, the infection pathway could only be speculated. For example, 25 of the deceased came from nursing homes or residential and care facilities for the disabled, in which other residents had been diagnosed with SARS-CoV-2 infections previously. In two cases, the infection was presumably transmitted in the hospital by fellow patients; in six cases, the infection was probably caused by travel to countries defined as risk areas at that time; and in one case, presumably by contact with family members at a family celebration.
A total of 76 cases (95%) were classified as COVID-19 deaths, corresponding to categories 1–3. Four deaths (5%) were unrelated to SARS-CoV-2 infection. In these four deaths, autopsy and radiological findings suggestive of COVID-19 were missing. The causes of death in these four cases were pericardial tamponade as a complication of myocardial infarction (case 39), sepsis secondary to necrotizing fasciitis (case 48), and two sudden cardiac deaths due to severe heart disease (CHD, dilated cardiomyopathy; cases 23 and 32). In three of these four cases, SARS-CoV-2 infection had been diagnosed postmortem by nasopharyngeal swabs. Typical COVID-19 symptoms such as cough, sore throat, impaired taste or smell, or flu-like infection were not known in any of the four cases antemortem. However, one of the deceased (case 23) was found to have an elevated body temperature postmortem.
Fifty-seven cases (71%) corresponded to category 1. In all these cases, pneumonia, with or without evidence of sepsis, was found to be the cause of death. Also in category 2, pneumonia was present in all 10 cases (25%). In seven of these 10 cases, however, a fulminant pulmonary artery embolism was fatal, and in one case each, aortic valve endocarditis, septic encephalopathy, and hepatorenal failure secondary to liver cirrhosis were contributory causes of death. A total of eight cases (10%) were classified in category 3 in which a competing cause of death is also considered in addition to COVID-19 (e.g. aspiration pneumonia, pronounced emphysema without evidence of pneumonia, or acute bronchitis). In these cases, a relation with SARS-CoV-2 infection can certainly be discussed critically.
In addition to the eight fatal fulminant pulmonary artery embolisms, peripheral pulmonary artery embolisms were found in nine other cases—a total of 17 cases (21%) altogether. In each of these deaths as well as in fifteen others (in total 32 cases, 40%), thrombi were found in the deep veins of the lower extremities. Of these deaths, 21 were male and 11 female (ratio 1.9:1). The male deceased also showed thrombi in the prostatic venous plexus in 15 cases and in the veins of the oesophagus in one case. The average BMI of these 32 deceased was 28.5 kg/m2. The places of death were hospitals (12 cases in the ICU, seven cases in the normal ward, one case in the emergency room), nursing homes (six cases), and own homes (four cases).
For the first 30 deceased autopsied, combined naso- and oropharyngeal swabs and swabs of the lung tissue were taken at the time of the dissection. In all these 30 cases, a SARS-CoV-2 infection could be diagnosed by PCR postmortem. The maximum PMI in these cases was 12 days.
Macroscopic lung findings revealed a broad spectrum of changes, often overlaid by chronic diseases such as chronic bronchitis and emphysema. It seems to be typical for COVID-19 pneumonia that the lungs are very large and heavy due to retained fluid (Fig. 1a). The mean combined lung weight was 1,610 g. Standard combined lung weights are 639 g (female) and 840 g (male) [9, 10]. The lung surfaces sometimes showed signs of pleurisy (Fig. 1d). Typically, a mosaic-like pattern of pale fields and slightly protruding dark purple sections with prominent capillary drawing has been seen (Fig. 1b). On the cut surfaces, the affected lung sections were either ubiquitously dark red or also alternately faded (Fig. 1c). The tissue was diffusely solidified but at the same time fragile. However, in some of the cases, lung changes seen in COVID-19-associated deaths appeared as a purulent respiratory tract infection with abscessed bronchopneumonia. In these cases, the typical macroscopic signs of acute respiratory distress syndrome were not very pronounced or were absent altogether.
Histologically, 8 cases (only the first 12 cases have been evaluated so far) showed diffuse alveolar damage (DAD) with activated type II pneumocytes, fibroblasts, protein-rich exudate, and hyaline membranes (Fig. 2a and d). In advanced stages, squamous metaplasia and fibrosis occurred (Fig. 2b). In some cases, giant cells and megakaryocytes appeared. The small pulmonary arteries often showed a pronounced infiltrate of lymphocytes and plasma cells, whereby the endothelia were not reactively altered in the sense of vasculitis (Fig. 2c). However, there were four cases in which the picture of a granulocyte-dominated focal confluent bronchopneumonia was dominant (Fig. 2e). In these cases, there was often chronic, purulent exacerbated bronchitis. Mixed forms of DAD and purulent pneumonia also occurred in different stages of organisation. Some patients had advanced stages of emphysema with destruction of alveolar septae, fibrosis, and lymphocytic infiltrates.
Other organs often showed signs of chronic diseases, such as scarring in the myocardium, arterio-arteriolosclerosis of the kidneys, and congestion of the liver and spleen. A small lymphocytic infiltrate in the right ventricle of the heart was present in one case as a sign of myocarditis (case 4).
Mild to pronounced lymphocytic pharyngitis was found in 7 of the 8 cases examined (Fig. 2f). Shock changes in the liver, kidneys, or intestine were found in half of the cases. The veins of the lower extremities showed no pathological changes, except for varying degrees of phlebosclerosis and fresh thrombi, and in particular, no signs of phlebitis.