Postmortem lung biopsies from four patients with COVID-19 at a tertiary hospital in Cape Town, South Africa

Abstract

Background. An outbreak of a novel coronavirus in China in late 2019 has resulted in a global pandemic. The virus (SARS-CoV-2) causes a severe acute respiratory syndrome and had been responsible for >14 000 deaths in South Africa (SA) at the time of writing, 30 August 2020. Autopsies in our setting have not been prioritised owing to the infective risks for staff, resulting in a lack of information on the histopathology of the disease in the SA setting. Postmortem biopsies are relatively quick and easy to perform and reduce the infective risk posed by full autopsies. Objectives. To determine whether postmortem biopsies of lung tissue could be used to determine cause of death in lieu of full autopsies in patients dying from COVID-19. Methods. We performed postmortem biopsies of lung tissue on 4 patients with SARS-CoV-2 confirmed by reverse transcriptase polymerase chain reaction who died in the Tygerberg Hospital (Cape Town, SA) intensive care unit (ICU) in June - July 2020, in order to determine their cause of death. The biopsies were performed in the ICU with the necessary personal protective equipment within 2 hours after death. Clinical information was obtained from the hospital records and the histopathology was reviewed by two consultant histopathologists. Microbiology and electron microscopy were also performed on this tissue. Results. All 4 patients were aged >50 years and had multiple comorbidities. Pulmonary pathology was present in only 3 cases, and the findings were surprisingly heterogeneous. One case demonstrated several findings including diffuse alveolar damage, extensive fibrin thrombi in pulmonary arteries with pulmonary infarction, organising pneumonia and bronchopneumonia. Other findings included type 2 pneumocyte hyperplasia, intra-alveolar macrophages and squamous metaplasia. An organising pneumonia was present in 2 other cases, although these findings were not deemed to be severe enough to be the cause of death. Fibrin thrombi were present in pulmonary arteries of 3 cases. One case showed no significant acute pulmonary pathology. The cause of death could only be determined in 1 case. Conclusions. The pulmonary findings we observed are in keeping with those described in the international literature. However, the pathology was surprisingly heterogeneous between cases, and was only deemed severe enough to be the cause of death in 1 of 4 cases. While lung-targeted, standardised postmortem biopsies may be safe, easy to perform and provide useful insights into the disease, they are not suitable to replace full autopsies in determining cause of death.