We submitted a COVID-19-related paper recently, reporting a longitudinal proteomic investigation of diverse serological patterns in COVID-19 (link to the paper: medRxiv). It is not a paper with fancy analysis nor sophisticated modeling per se, yet the journey of shaping up the research work is quite interesting.
Briefly, during COVID-19, we found that some patients did not secret serum IgG or IgM (seronegative) whereas others secreted drastically high levels of serum IgM. We thus used clinical indicators and serum proteomics to investigate the underlying mechanisms.
These unusual clinical phenomena were observed over two years ago, during an offline meeting with my clinical collaborators from Taizhou Hospital. This is of particular interest to us, among other research ideas, in that although these seronegative patients were specially attended to during hospitalization due to negative results of serological tests, all of them exhibited mild symptoms. They did not require intense medical treatment or oxygen inhalation, and had a relatively shorter length of hospitalization time. We decided to initiate a project studying the molecular mechanisms behind the serological diversity, and thus fetched the archived serum samples for analysis.
In the meantime, more and more papers were coming out reporting COVID-19 serology. We found that serological negativity was a pervasive phenomenon, although not specifically reasoned in these papers. Two papers attributed this to the limitation of sensitivity regarding the antibody detection kit, or a lack of continuous sampling, which failed to cover the seropositive stage. The overly high expression of IgM or IgG, comparatively, was less reported and discussed. Most papers ended up calculating a positive rate for antibodies, which lost the quantitative information.
In our data, we conducted a continuous serology sampling during the 70 days since COVID-19 onset. We found that the seronegative patients remained negative in both IgM and IgM throughout the period, and the patients with overly high IgM expression had at least two data points with 20-fold higher IgM reads than the average. Also, Tukey’s test showed that their antibody reads were the low and high outliers in the group, respectively. Additionally, their clinical indicators were significantly different in expression compared to others, whether on admission or during disease progression. The above evidence suggests that there exist unique mechanisms for the regulation of their antibody expression.
What further supported our idea was the one-year revisit data of these patients, wherein seronegative patients remained negative in both IgM and IgG, whereas others were mostly IgG positive. The two-year revisit data was interfered by the vaccination of these recovered patients, but we found the seronegative patients turned positive whereas the IgM overly high phenomenon was not observed, suggesting that the serological diversity might be distinctly related to SARS-CoV-2 virus instead of any external perturbation.
With opinions on herd immunity and frequent outbreaks of the COVID-19 pandemic, there emerged debates on the means of large-scale COVID-19 screening. Especially, during this April’s COVID-19 pandemic in Shanghai, China, a rising voice advocates the massive prioritization of serological tests to evaluate antibody expression levels, which may represent the population resistance towards SARS-CoV-2, to complement the current viral nucleic acid tests. Our work provided some mechanistic evidence regarding the unexpected serological behaviors, and suggested that a more careful interpretation of serological results should be considered.
Note: The opionions in the blog are on my own.
Left, first figure draft on 20200914; Right, figure draft on 20220701.