Issue #117: Persistence of SARS-CoV-2 and Long COVID - The spike in the blood

Written by Nobel Laureate Professor Peter Doherty

Looking at Long COVID (LC) or, as many US investigators prefer, PASC (post-acute sequelae of COVID-19), it’s obvious that a subset of those suffering persistent clinical compromise were hospitalised with severe symptoms during the acute phase of the infection (#106 and #110). Indeed, time spent in an ICU is a predictor of continuing problems, with that post intensive care syndrome (PICS) manifestation not being restricted to those suffering SARS-CoV-2 infection. 

For patients who did not survive the COVID-19 ICU experience, an exhaustively analysed mortem series (#116) showed that the virus goes everywhere. When it comes to those who were able to return home after time in a COVID ICU and recover (to a greater or lesser extent), it’s not unreasonable to speculate that they also experienced widespread virus dissemination and organ damage. Whether any LC/PASC/PICS symptoms in these patients reflect an active process due to virus persistence - or are simply a consequence of permanent virus-associated damage occurring in the acute clinical phase - is unclear. If the latter is the case, the main clinical options are likely to be symptom-directed treatments and rehabilitation medicine. 

But what of patients who may experience a relatively mild COVID-19 clinical course, or symptoms that do not merit hospitalisation, yet still develop LC/PASC of varying duration and severity? If virus persistence is characteristic of these cases, that raises the possibility that treatment, perhaps with vaccines (#96 and #109), or with antiviral drugs (#66) or virus-specific monoclonal antibodies (#76) could promote recovery. 

A very recent paper, that, while from leading institutions (Harvard, MIT), is still at the preprint stage, indicates that this may indeed be a possibility. The manuscript is a bit unclear in places and will obviously benefit from the comments of reviewers (every publishing scientist has that experience!), but their main message is so relevant to the discussion we’ve been having recently (#111-116) that I’ll push on regardless.

Using very sensitive protein detection assays to look for the SARS-CoV-2 spike, the S1 fragment (the bit that binds to the ACE-2 molecule) or the viral nucleocapsid (N), the investigators analysed one or more plasma samples from 63 individuals who had experienced COVID-19. All had tested positive by PCR during the acute phase of the infection. Of these, 31 of the 37 (30 were women) that developed LC/PASC (only two were hospitalised and both had an ICU stay) were followed for up to 12 months, while the remainder (10 out of 26 were in ICU and seven were intubated) were studied over five months. Using the author’s categorisation, we’ll refer to these two groups as the ‘PASC’ and ‘COVID-19’ series. 

For the COVID-19 set, many of whom had a more severe clinical experience during the acute phase, they found high levels of S1 (with no spike), and often N, in blood over the first week after diagnosis, but not thereafter. With the PASC cohort, 30 of whom provided at least one or two plasma samples while 12 were sampled repeatedly for longitudinal analysis, they were able to detect S1, spike or N in 65% of cases at any given time point, with spike (60%) being the most common out to 12 months. About 20% were positive for S1, while N was found in only one patient, but on multiple occasions. The authors speculated that N is less readily detected because it is more rapidly broken down (hydrolysed), though finding spike and N (which is an internal virion protein) together could indicate that infectious virus is indeed being made.

If you go to the paper and look at the results shown in Figures 2 and S1, you will see the very clear difference between the PASC and COVID-19 groups. The overall conclusion by these researchers is that there is clear evidence for the maintenance of SARS-CoV-2 reservoirs in PASC patients, though it could also be the case that we are seeing the persistent production of defective virus. In the future, it would be great to see more comprehensive, longitudinal studies that look both at antigen levels and the SARS-CoV-2 specific antibody (IgG), B cell and T cells responses. Hopefully, some research group will do that.

For the vaccine hesitant, these investigators earlier looked at 13 clinically normal subjects (some of whom may have had subclinical COVID-19) who were primed and boosted with 100 g of the Moderna mRNA vaccine. Spike protein was detectable at a low level in 3 out of 13 patients after the first dose while S1 was found in most up to day 9 when virus-specific antibody levels (IgG) started to shoot up (see Figure 1). No viral protein was detected after the booster shot, presumably because the IgG was taking it out. Nothing remarkable to see here: this is exactly what any immunologist would expect for a vaccine! 

Some of the PASC patients in the PASC/COVID comparison were also given a second or third dose of (presumably) an mRNA vaccine (as shown in Figure 2) during the course of their illness, but there was no clear indication that this led to the elimination of the spike protein from the blood.  What is obvious, however, is that virus (or viral protein) persistence is likely to be a factor in at least some LC/PASC cases that develop (and even progress) following a relatively mild initial experience of COVID-19.