Issue #116: Persistence of SARS-CoV-2 and Long COVID - Looking at a post-mortem series

Written by Nobel Laureate Professor Peter Doherty

From the early days of COVID-19, it’s been clear – from sequential PCR checked by gene sequencing – that a few people can be long-term ‘carriers ‘of (at least) SARS-CoV-2 RNA. A 2020 study from Wuhan identified 38 individuals (mostly aged 65+), from 50,000 hospitalised patients, and found viral RNA in oral swabs and sputum for 92 (58-118) days. Held under observation at the Wuhan Pulmonary Hospital, 29 of the 38 were asymptomatic and eight others had relatively mild symptoms, with all showing evidence of viral pneumonia in chest CT scans. Since then, with the regular replacement of SARS-CoV-2 variants by more recently evolving strains, it’s become obvious that any such potential ‘carriers’ are not playing a significant part in maintaining SARS-CoV-2 in nature.

Last week (#115), looking from a Long COVID (LC) perspective, we discussed more recent data on SARS-CoV-2 RNA shedding for people who had generally experienced a relatively mild initial course of COVID-19. Focusing on the ‘inside/outside’ environments of our respiratory and gastrointestinal (GI) tracts, continued shedding in faecal samples (which remained PCR+ for longer than swabs from the nasopharynx) correlated with both LC GI tract symptoms and with the presence of viral protein in the enterocytes of gut wall.

Many of the symptoms of LC seem, though, to involve major internal organs, like the heart, brain, and muscle, suggesting that, if components of SARS-CoV-2 are driving continued clinical impairment, the relevant viral RNA and/or protein must be distributed more broadly through our bodies, not just at the ‘immune surveillance frontier’ of mucosal surfaces. Though imaging of live patients can be incredibly sophisticated for many applications, the only way to look intensively for the location and distribution of viral components in the ‘inside/insides’ of us (#115) is to analyse samples obtained by biopsy (an obviously limited strategy) or at necropsy. Just published as a preprint (it is yet to be peer reviewed), the findings from an intensively-studied autopsy series are incredibly informative.

The analysis describes 44 cases from SARS-CoV-2 positive patients who succumbed as late as seven months after their initial infection with SARS-CoV-2. The cause of death was listed as COVID-19 in 39, with the others dying ‘with COVID’ from some co-morbidity. A primary focus was intensive microscopic examination of tissue (for example, Figures 2 and 3 in the paper), using immunohistochemistry (ICH) – staining with fluorochrome-labelled antibodies to identify viral protein and cell types – and in situ hybridisation (ISH) to show the location of viral RNA. Variants of the PCR technique were used with tissue extracts to quantify genomic RNA and sub-genomic RNA (sgRNA), that’s considered to be a marker of recent virus replication. The subjects were classed as early (n=17), mid (n=13, day 15-30) or late (n=14, from day 31) with the analysis of persistence being from the late series.

Reflecting the systemic (blood-borne) nature of this infection, infectious virus was isolated from multiple tissues of early cases up to day 7, though only two showed the presence of viral RNA in plasma indicating that blood contamination was not the source. Otherwise, plasma was generally negative and, using the combination of ICH and ISH, viral RNA and protein were localised to individual cells in the various anatomical locations. In all, it was found in 36 different cell types. When it came to the detection of viral RNA in tissues, every case was positive while sgRNA was found across all tissues in the early cases, and for at least one sampling site for 61.5% of the mid and 42.9% late series. The calculated overall burden of SARS-CoV-2 RNA dropped by a log or more across tissue categories for the ‘mid’ cases, and decreased further in the ‘late’ subjects, though the results for a few individuals remained high.

Looking at the heart, for example, spike RNA was found in cardiac myocytes, the endothelium and smooth muscle of blood vessels, fibroblasts in the pericardium and the intima of the aorta. Viral RNA was also present in a variety of cell types within the ‘immunologically-privileged’ testis. In the brain, spike RNA was found in neurons (including a spectrum of types in the cerebellum), glia, ependyma and the blood vessels. Apologies for the technical nomenclature, but the basic point is that components (at least) of the SARS-CoV-2 virus can be found in a diversity of cell types through the main body organs of people with severe COVID-19, including those who survived for months after the acute stage of the infection.

Obviously, these were very sick people, with 81.8% having been intubated in an ICU, 22.7% needing ECMO (oxygenating the blood by passing it through a machine) and 40.9 % receiving renal replacement therapy. The experienced pathologists who analysed these samples were surprised at the relative lack of invading inflammatory cells, like T lymphocytes and monocytes, in sites where there looked to be a lot of virus. Was the fact that these people died reflective of a failed cellular immune response?

How useful are these findings for the understanding of LC? They likely tell us a lot about the type of LC we see in ICU survivors. But analysing the situation regarding virus persistence for those who develop LC after a mild initial course is, for obvious reasons, less accessible.