Plasma-based antigen persistence in the post-acute phase of SARS-CoV-2 infection

BACKGROUND

Early in the COVID-19 pandemic, there were case reports of prolonged detection of SARS-CoV-2 RNA, well past the period of acute symptoms, in surface samples from the nasopharynx and gastrointestinal tract, two sites known to be involved in primary infection. Systematic investigation, however, revealed RNA detection after 28 days following symptom onset in immunocompetent hosts was uncommon and rarely documented to exceed 90 days.^1–7 More recently, there has been detection of either SARS-CoV-2 RNA or protein in other sites, such as blood or gastrointestinal tract tissue, at durations of three months or more following acute infection.⁸ These findings suggest a phenomenon beyond that of the residual tail of primary viral infection; they suggest a persistent SARS-CoV-2 reservoir. Given the sheer number of persons who have been infected with SARS-CoV-2, a persistent reservoir of virus — even if clinically relevant in a small percentage of persons — could have substantial implications at the population level.

While intriguing, the initial data supporting persistent SARS-CoV-2, in some form or location, are short of definitive. Use of small and non-representative sample sizes, unclear documentation of complicating vaccination and reinfection histories, and, most importantly, lack of a sizeable true-negative comparator group to assess assay specificity all threaten conclusions about the existence, frequency, and duration of viral persistence.^9–16 To address these limitations, we evaluated the presence of SARS-CoV-2 antigens in plasma in a large, well-characterized group of persons at several time points in the first 14 months of the post-acute phase of SARS-CoV-2 infection, most of whom were studied prior to widespread vaccination or reinfection. To understand the specificity of our findings, we compared them to a large number of persons studied prior to 2020, who, by definition, were SARS-CoV-2-uninfected. Finally, we evaluated several host and acute COVID-19-related factors for their influence on persistent SARS-CoV-2 antigenemia.

METHODS

RESULTS

SARS-CoV-2 antigen detection in pandemic era compared to pre-pandemic era

Positivity in at least one SARS-CoV-2 antigen assay was present in the plasma of 5 pre-pandemic participants (2%; 95% confidence interval (CI): 0.65% to 4.6%). Given the definitional absence of target in these specimens, these 5 instances are considered false positive, thus establishing any-antigen assay specificity to be 98% (95% CI: 95% to 99%). For the individual antigens, spike was detected in 3 (1.2%) participants, S1 in 3 (1.2%), and N in 2 (0.80%). The 5 participants were a 20-to-29-year-old White woman (S1 alone [34.67 pg/mL]), 50-to-59-year-old Asian man (spike alone [83.46 pg/mL]), 40-to-49-year-old White man (spike alone [609.96 pg/mL]), 30-to-49-year-old Asian man (S1 [285.31 pg/mL] and N [649.28 pg/mL]), and 40-to-49-year-old White woman (spike [646.02 pg/mL], S1 [115.89 pg/mL], and N [5716.32 pg/mL]). Of the 5, 2 (40%) were HIV-seropositive.

Of the 660 pandemic-era plasma specimens tested, 61 (9.2%) representing 42 unique participants (25% of the group) had one or more detectable SARS-CoV-2 antigens (Figure 1). The most commonly detected antigen was spike (n=33, 5.0%), followed by S1 (n=15, 2.3%) and N (n=15, 2.3%). In most instances (59/61, 97%) only a single antigen was detected; one specimen was positive for S1 and N and a second was positive for spike and N. Of those with detectable antigen, the median (IQR) concentration was 27.7 pg/mL (IQR 20.5 to 33.7) for spike, 31.2 pg/mL (IQR 20.5 to 193.0) for S1, and 23.6 pg/mL (IQR 6.46 to 62.0) for N.

• Download figure
• Open in new tab

Figure 1.

Prevalence of SARS-CoV-2 antigen positivity in plasma among participants in the post-acute phase of COVID-19 in comparison to pre-pandemic participants. P-values represent chi-square and Fisher’s 2-sided exact test as appropriate. (a) Prevalence of any SARS-CoV-2 antigen positivity (Spike, S1 or nucleocapsid); (b) Spike antigen prevalence; (c) Nucleocapsid antigen presence; and (d) S1 antigen presence.

Compared to the pre-pandemic group, detection of any SARS-CoV-2 antigen was significantly more frequent among pandemic era participants at all three time periods that we evaluated in the post-acute phase of infection. The absolute difference in antigen prevalence was +11% (95% CI: +5.0% to +16%) at 3.0-6.0 months post-onset of COVID-19; +8.7% (95% CI: +3.1% to +14%) at 6.1 to 10.0 months; and +5.4% (95% CI: +0.42% to +10%) at 10.1-14 months (Figure 2a; Supplemental Table 1). Regarding the individual antigens, significant differences between pandemic-era and pre-pandemic-era participants were observed for spike protein for up to 10 months and for N in the first 6 months after infection (Figure 2b-d).

View this table:

• View inline
• View popup
• Download powerpoint

Supplemental Table 1.

Differences in prevalence of SARS-CoV-2 antigen positivity in plasma among participants in the post-acute phase of COVID-19 in comparison to pre-pandemic participants. P-values represent chi-square and Fisher’s 2-sided exact test as appropriate.

• Download figure
• Open in new tab

Figure 2.

Quantitative measurement of SARS-CoV-2 spike, S1, and nucleocapsid antigens in plasma among all pandemic-era participants during the post-acute phase of SARS-CoV-2 infection. Dashed lines indicate limit of quantification for each antigen. Y-axis refers to log-transformed concentration of antigen in picograms per mL.

Profiles of SARS-CoV-2 antigen positivity over time within individual pandemic-era participants

Of 159 pandemic-era participants who had multiple timepoints studied, 29 (18%) had antigen detected at a single post-acute timepoint, 10 (6.3%) had antigen detected at two post-acute timepoints, and one (0.63%) had antigen detected at three, four, and five post-acute timepoints, respectively (Figure 3). Most timepoints at which antigen was detected (51/61, 84%) occurred before the participant had ever received a SARS-CoV-2 vaccine. There were five instances in which antigen was detected within three weeks of a SARS-CoV-2 vaccine dose (three for S1, one for Spike, and one for N).

• Download figure
• Open in new tab

Figure 3.

Individual participant-level profiles of quantitative measurement of SARS-CoV-2 spike, S1, and nucleocapsid antigens in plasma during the post-acute phase of SARS-CoV-2 infection, limited to participants with at least one positive antigen. Blue indicates spike, green nucleocapsid, and red S1. Horizontal dotted lines represent the assay limit of detection for each antigen. Vertical dotted lines indicate receipt of SARS-CoV-2 vaccine. Y-axis refers to log-transformed concentration of antigen in picograms per mL.

DISCUSSION

Persistent symptoms among some persons who develop COVID-19 has led to the hypothesis that SARS-CoV-2 may, in some form or location, persist in the human host for long periods following acute infection.¹⁹ This hypothesis has motivated the search for SARS-CoV-2 far removed in time from acute infection and in sites disparate from those of acute infection.⁸ Overcoming many of the limitations of the initial work in this regard, we found evidence of SARS-CoV-2 in plasma for up to 14 months following SARS-CoV-2 infection and that this persistence is influenced by the events of acute infection.

Because our measurement of SARS-CoV-2 is via immunoreactivity, it cannot be assumed that every signal above background is specific for antigen from natural SARS-CoV-2 infection. Antigens from related pathogens or the host can theoretically cross-react. This differs from detection of nucleic acid for which specificity is more certain and more easily directly assessed by sequencing. Therefore, understanding the specificity of our Simoa® assay is critical for interpreting findings in SARS-CoV-2-infected persons, and this requires a true-negative group large enough to precisely estimate low-frequency false positivity. Using a diverse group of 250 adults sampled prior to the introduction of SARS-CoV-2 in humans, we found 2% false positivity in the assay. Of note, the quantitative values for the false positive specimens were well above limits of detection but not overtly related to any available demographic characteristic. This 98% specificity is high but not complete, meaning that caution would be needed if interpreting individual-level results in patients. Although not formally false positivity, there is also concern that vaccination against SARS-CoV-2 or recent reinfections could cloud interpretation of positive assay results.⁸ We mitigated this by studying specimens from an era that was largely prior to either of these occurrences and by detailed characterization of our participants.

Our inferences are consistent with most but not all prior smaller studies investigating SARS-CoV-2 persistence three or more months following acute infection. In plasma, Swank et al. (n=40 SARS-CoV-2-infected participants versus n=45 uninfected),¹³ Schultheiß et al. (n=29 versus n=2),¹⁵ and Craddock et al. (n=47 versus n=15)¹⁶ each found excess SARS-CoV-2 antigen prevalence in the post-acute period compared to uninfected persons, but Kanberg et al. (n=31 versus n=17) did not.¹⁴ Studies of tissue have featured even smaller numbers of participants,⁸ the exception being an investigation of 46 participants undergoing colonoscopy for inflammatory bowel disease in whom 70% had detectable SARS-CoV-2 RNA or antigen.¹² The proportion of antigen positivity in our study was lower than most of the prior reports,^15,16 including one using our assay.¹³ This may be explained by other studies being enriched with highly symptomatic patients recruited from Long COVID clinics whereas our population was consecutive volunteers who were interested in a variety of long-term aspects related to COVID-19. Alternatively, our work and the cited studies have used three different never-compared antigen assays that may not be interchangeable. In addition to direct detection of SARS-CoV-2 components in the post-acute phase of infection, evolving B cell immunity over time also indirectly suggests that some aspect of SARS-CoV-2 is persistent.⁹ Finally, regarding biologic plausibility, there is ample evidence of persistence of feline coronaviruses in their natural host.^20,21

That persons hospitalized during acute COVID-19 had a significantly higher probability of persistent SARS-CoV-2 antigenemia, an observation also described by others,¹⁶ suggests the influence of the acute phase of infection in establishing the persistent SARS-CoV-2 reservoir. This was bolstered by the finding, although not meeting conventional statistical significance, that those reporting the most severe overall health during acute infection among the non-hospitalized also had substantially higher frequency of post-acute antigenemia. Coupled with the recent documentation of replication-competent virus in blood during acute infection,²² our findings suggest that SARS-CoV-2 might seed distal sites through the bloodstream and establish protected reservoirs in some sites. Alternatively, clinically more severe acute infection may simply be a marker of higher inoculum in the known sites of primary infection, which then have a greater chance of evading subsequent immune clearance. Our findings, however, provide no direct evidence regarding the persistent presence of replication-competent or even transcriptionally active virus. Yet, finding of circulating antigen so far in time from acute infection at least suggests some source of antigen production to counteract clearance. Whatever the mechanism for establishing a reservoir, our findings suggest an explanation for the protective effect of SARS-CoV-2 vaccination and antiviral therapy during acute infection against the occurrence of post-acute sequelae of COVID-19 (PASC).^23–25

As in a prior report using our assay,¹³ antigen detection was intermittent over time in individuals in whom it was ever present. To what extent this represents true biologic variability versus lack of assay reproducibility is unknown. Formally, if true biologic variability, this could stem from variability in antigen production, release from reservoirs into the bloodstream, or clearance. A more sensitive assay that could interrogate larger plasma volumes, such as one being developed by our team,²⁶ might provide substantial insights on these questions.

Our work has several limitations. First, our pandemic era group was a convenience sample, enriched to some unknown degree by those experiencing symptoms. Thus, the approximately 5% to 10% excess prevalence of antigenemia that we observed cannot be taken as a population-based estimate for all individuals experiencing SARS-CoV-2 infection during the period of our study. However, we suspect identifying a more representative group, replete with available blood specimens, will be highly challenging. Second, our participants were mainly infected with ancestral SARS-CoV-2 strains without the benefit of prior vaccination or effective anti-viral agents. It is thus unclear how our findings will generalize to contemporary infections. Finally, while our participants did not have known or clinically suspected reinfections prior to antigen detection, they were not systematically assessed for asymptomatic SARS-CoV-2 infection. It is therefore possible that some fraction of the antigenemia is from recent re-infection instead of the distant primary infection.

In summary, our data provide among the strongest evidence to date that SARS-CoV-2 should be added to the list of RNA viruses whose components may persist beyond the period of acute illness. Given the scale of the pandemic and knowledge that SARS-CoV-2 spike is highly immunogenic,^27–29 there are now urgent questions regarding whether persistent SARS-CoV-2 is causally related to either the chronic symptoms of PASC (e.g., fatigue, pain, and cognitive difficulty) or discrete incident complications (e.g., cardiovascular events). Addressing this agenda will require observational as well as concurrent experimental studies that target inhibition of viral component production, neutralization or clearance of existing antigen, and/or modulation of the immune response to these antigens. Such studies are now underway (NCT05877507, NCT05595369, NCT05576662, NCT05668091, and NCT06161688).

FOOTNOTES

Author Contributions

Designed the study: MJP, SGD, DRW, JNM

Designed the cohort: MJP, PYH, MSD, TJH, JDK, SGD, JNM

Collected biospecimens and clinical data: MJP, RH, ML, CCS, AC, AZ, ML, BH, KA.

Performed the assays: ZS, EB, YS in the laboratory of DRW.

Managed the data: SL, TD, BV.

Analyzed the data: MJP, SAG, DVG.

Wrote the manuscript: MJP, SGD, JDK, DRW, JNM.

Edited the manuscript: All authors.

Approved the manuscript: All authors.

Funding

This work was supported by funding from the PolyBio Research Foundation to support the LIINC Clinical Core, as well as support from NIH/NIAID 3R01AI141003-03S1, NIH/NIAID R01AI158013, and NIH/NIAID K23AI134327.

Conflicts of Interest

MJP reports consulting for Gilead Sciences and AstraZeneca, outside the submitted work. DRW has a financial interest in Quanterix Corporation, a company that develops an ultra-sensitive digital immunoassay platform. He is an inventor of the Simoa technology, a founder of the company, and also serves on its Board of Directors. Dr. Walt’s interests were reviewed and are managed by Mass General Brigham and Harvard University in accordance with their conflict-of-interest policies.