Asymptomatic coronavirus sufferers appear to lose detectable antibodies sooner than people who have exhibited COVID-19 symptoms, according to one of the biggest studies of its kind in Britain published on Tuesday.
The findings by Imperial College London and market research firm Ipsos Mori also suggest the loss of antibodies was slower in 18-24 year-olds compared to those aged 75 and over.
Overall, samples from hundreds of thousands of people across England between mid-June and late September showed the prevalence of virus antibodies fell by more than a quarter.
The research, commissioned by the British government and published Tuesday by Imperial, indicates people’s immune response to COVID-19 reduces over time following infection.
James Bethell, a junior health minister, called it “a critical piece of research, helping us to understand the nature of COVID-19 antibodies over time”.
But scientists involved cautioned that a great deal remains unknown about people’s long-term antibody response to the virus.
“It remains unclear what level of immunity antibodies provide, or for how long this immunity lasts,” said Paul Elliott, of Imperial’s School of Public Health.
The study involved 365,000 randomly-selected adults administering at home three rounds of finger prick tests for coronavirus antibodies between June 20 and September 28.
The results showed the number of people with antibodies fell by 26.5 percent over the approximate three-month period.
Scaled up to a nationwide level, it meant the proportion of the English population with antibodies dropped from 6.0 percent to 4.4 percent, according to the study.
The decline coincided with the prevalence of the virus falling dramatically across England – and the rest of Britain – following a months-long national shutdown which was eased over the summer.
However, the research found the number of health care workers testing positive for antibodies did not change over time, potentially reflecting repeated, or higher initial, exposure to the virus.
“This very large study has shown that the proportion of people with detectable antibodies is falling over time,” said Helen Ward, one of the lead authors.
“We don’t yet know whether this will leave these people at risk of reinfection with the virus that causes COVID-19, but it is essential that everyone continues to follow guidance to reduce the risk to themselves and others.”
Antibody Production of Asymptomatic to Pneumonic COVID-19 Patients
Antibody production of asymptomatic to pneumonic COVID-19 patients are presented in Figure 1 and Table 2. The neutralizing antibody was detected in 91.4% (64/70) of the enrolled COVID-19 patients.
All the patients in the pneumonia group produced the neutralizing antibody (100%). Production of the neutralizing antibody was observed in 93.9% of the mild symptomatic group and 80.0% of the asymptomatic/anosmia group. Although the proportion of neutralizing antibody production was lower in the milder groups, statistical significance was not noticed (p = 0.079).
In the comparison of high MN titers between groups, all the patients in the pneumonia group produced the neutralizing antibody with MN titer ≥ 1:80, while 36.7% of the mild symptomatic group and 20.0% of the asymptomatic/anosmia group showed MN titer ≥ 1:80 (p < 0.001).
Different proportions of high MN titers between severity groups and statistical significance were also noticed in the comparison of the five severity groups (p < 0.001, Table S2).
Table 2 – Neutralizing antibody production and serologic tests for asymptomatic to severe COVID-19 patients.
n = 15
|Mild Symptomatic Group|
n = 49
n = 18 *
|Positive, titer ≥1:10||12 (80.0%)||46 (93.9%)||18 (100%)||0.079|
|High MN titer, ≥1:80||3 (20.0%)||18 (36.7%)||18 (100%)||<0.001|
|Positive, COI ≥1.1||15 (100%)||48 (98.0%)||18 (100%)||0.711|
|High COI value, ≥15.0||11 (73.3%)||44 (89.8%)||18 (100%)||0.049|
|Positive, OD ≥ 1.0||15 (100%)||47 (95.9%)||18 (100%)||0.501|
|High OD ratio, ≥ 3.0||8 (53.3%)||32 (65.3%)||18 (100%)||0.006|
Data are expressed as the number (%) of sera. * Eighteen sera from six patients. Abbreviations: COVID-19, coronavirus disease 2019; MN, microneutralization; FIA, fluorescence immunoassay; IgG, immunoglobulin G; COI, cut-off index; ELISA, enzyme-linked immunosorbent assay; OD, optical density.
With the FIA and ELISA methods, the antibody against SARS-CoV-2 in the asymptomatic/anosmia and mild symptomatic groups could be detected with a higher proportion than the MN test. All the patients in the asymptomatic/anosmia group were positive for FIA IgG and ELISA.
Only one or two patients in the mild symptomatic group were negative for FIA IgG and ELISA, respectively. For the FIA IgG test, all patients in the pneumonia group exhibited a COI value ≥ 15.0, while 89.8% of the mild symptomatic group and 73.3% of the asymptomatic/anosmia group showed a COI value ≥ 15.0 (p = 0.049).
For the ELISA test, all patients in the pneumonia group showed an OD ratio ≥ 3.0, while 65.3% of the mild symptomatic group and 53.3% of the asymptomatic/anosmia group showed an OD ratio ≥ 3.0 (p = 0.006).
Different proportions of high FIA IgG COI values and ELISA OD ratios between severity groups and statistical significance were also noticed in the comparison of the five severity groups (p = 0.039 and p = 0.034, respectively; Table S2).
Based on the test results of the present analysis, sensitivity was 98.8% for FIA IgG and 97.6% for ELISA. Although a false positive test result was not observed, the number of negative control samples was limited to calculate specificity.
The FIA IgM antibody was positive in three sera of the asymptomatic/anosmia and mild symptomatic patients (3/64, 4.7%), while nine sera (9/18 sera, 50.0%) from three pneumonic patients (3/6 patients, 50.0%) were positive for IgM (Figure S2).
Since the positive proportions of FIA IgM were relatively low, we additionally evaluated the serial serum specimens of two COVID-19 patients with pneumonia (Figure S3). Patient A was a 46-year-old male with underlying diabetes mellitus. He was admitted on day 4 of illness and all antibody tests were negative by day 5. On day 9 of illness, all IgM, IgG, and total antibodies converted to be positive.
The COI value of FIA IgM increased until discharge (day 19 of illness), while the COI value of IFA IgG and OD ratio of ELISA total antibody tests peaked on day 9 and did not show a further increment on the follow up specimens. Patient A developed pneumonia after admission and was treated with hydroxychloroquine, lopinavir/ritonavir, and inhaled ciclesonide.
He required oxygen support up to 5 L/min via nasal prolong. He could fully recover from illness without mechanical ventilator support and was discharged on day 20 of illness.
Patient B was a 66-year-old female with underlying diabetes mellitus. She was referred to our center on day 20 of illness due to pneumonia progression and received endotracheal intubation with mechanical ventilator support. All the antibody tests were positive from the day of referral, but IFA IgM COI values were much lower than those of Patient A measured on similar illness days.
She was treated with hydroxychloroquine, lopinavir/ritonavir, and intravenous corticosteroid. Her FIA IgM COI values showed a waning trend during illness, FIA IgG COI values remained high throughout the hospitalization, and OD ratio of ELISA total antibody tests showed an increasing trend. Patient B fully recovered from the illness and was discharged on day 57 of illness.
Since not all serial specimens of these patients underwent an MN test, the titers of MN titers were not presented in the figure. The MN titers of Patient A increased from day 9 of illness, and peaked on day 13 of illness (1:320). Patient B showed similar MN titers during hospitalization (1:80 or 1:160).
In the present serologic investigation, a neutralizing antibody was detected in mild COVID-19 patients, even in asymptomatic cases. However, not all the asymptomatic and mild COVID-19 patients could produce neutralizing antibodies, and high titer production was significantly lower in asymptomatic to mild patients, compared to that of moderate to severe patients with pneumonia.
Immunoassays including FIA and ELISA methods could detect past SARS-CoV-2 infections sensitively, and different productions of high FIA COI values and high ELISA OD ratios were also noticed among the severity groups. These findings may have important implications regarding the practical application of serologic tests as follows.
To cope with the current COVID-19 pandemic, a sero-prevalence investigation of a community is conducted to assess the proportion of subclinical infections and status for herd immunity [9,28,29]. This approach has three major assumptions. First, there would be a certain proportion of subclinical infections which may not be detected by a symptom-based, RT-PCR screening strategy.
Subclinical cases have been reported to be 20–40%, though the undetected proportion would vary depending on the outbreak situation [8,30]. Second, subclinical infections may produce detectable antibodies against SARS-CoV-2. For the case of the Middle East respiratory syndrome coronavirus (MERS-CoV), this assumption may not be applicable since none of the asymptomatic cases and 60% of the febrile patients without pneumonia produced detectable antibodies .
In the present investigation, it was noticed that asymptomatic and mild COVID-19 patients could produce a detectable level of antibodies against SARS-CoV-2.
One completely asymptomatic patient even showed high MN titer of 1:80, similar with pneumonic COVID-19 patients. These findings imply that asymptomatic COVID-19 patients are not likely to be at the simple carrier status.
Without provoking recognizable symptoms, immunologic interactions between SARS-CoV-2 and the human body might occur silently. The immunologic phenomenon in asymptomatic and mild symptomatic COVID-19 patients needs to be further evaluated.
Third, the produced antibody may protect the host from re-infection of SARS-CoV-2. However, the production of the neutralizing antibody in asymptomatic to mild febrile COVID-19 patients was apparently lower than patients with pneumonia.
Although immunoassay methods including FIA and ELISA could sensitively detect anti-SARS-CoV-2 antibodies, this detection did not necessarily mean neutralization activity, especially in mild cases. It is not clear whether neutralizing antibodies with low titers may protect hosts from re-infection of SARS-CoV 2 and how long these antibodies may persist.
For the proper assessment of herd-immunity and the establishment of an outbreak control strategy, further studies should be conducted for the protective capacity and persistence of neutralizing antibodies.
From the perspective of the sero-diagnosis of SARS-CoV-2 infection, there could be several clinical applications. For the epidemiologic investigation of an outbreak cluster, serologic studies can be used to detect the asymptomatic index case of a cluster .
The finding of the present study is that most asymptomatic and mild COVID-19 cases produced detectable antibodies which supports the application of serologic tests for epidemiologic tracing. Both immunoassay methods of FIA IgG and the ELISA total antibody test showed a high sensitivity in detecting past SARS-CoV-2 infection regardless of the severity of illness. Considering the mechanism of tests, other immunoassay methods with well validated performances could be used in sero-epidemiologic studies.
Meanwhile, based on the findings of the present analysis, negative neutralizing tests cannot exclude past infections since 20% of RT-PCR confirmed asymptomatic/anomia COVID-19 patients were negative for neutralizing tests. Although we did not calculate the specificity of the IFA and ELISA kits due to the limited number of negative control specimens, it was noticed that a cross-reactivity to convalescence sera from other respiratory virus infections was not observed.
This finding has a similar implication for the serologic evaluation for healthcare workers (HCWs) who are caring for COVID-19 patients. Due to the high infectivity of SARS-CoV-2 and the heavy work burden to HCWs during an outbreak situation, HCWs are considered a high-risk group for SARS-CoV-2 infection.
By detecting past asymptomatic infections of HCWs, the overall risk and protective measures for HCWs can be evaluated. Sero-prevalence studies published to date are presented in the introduction section.
Whether to confirm positive immunoassay tests by neutralization tests, and whether HCWs with neutralizing antibodies would be safe during additional SARS-CoV-2 exposure, are still unresolved issues and need further investigations.
Serology data for various clinical spectrums of COVID-19 patients would be background knowledge for preparing CP therapy. Clinical studies in patients with acute respiratory distress syndrome caused by SARS-CoV-2 have shown a therapeutic benefit from the use of CP [12,31].
However, during the MERS outbreak in South Korea, CP was donated from relatively mild MERS patients who recovered shortly, but half of the donated plasma did not have neutralization activity . Likewise, young and previously healthy COVID-19 patients are more likely to be in good general condition after illness and willing to donate CP, but they might not have a sufficient concentration of the neutralizing antibody.
Instead, pneumonic patients with moderate illness may be better candidates for CP donors, since they show comparable neutralization titers to patients with severe illness with the minimum MN titer of 1:80.
Although the number of pneumonic patients was limited in the present analysis, the previous serologic study by Perera et al. also exhibited that the production of the neutralizing antibody was favorable in pneumonic patients with moderate illness .
As measurement of the neutralizing antibody titer and preservation of collected blood product are hurdles for CP treatment, hyperimmune immunoglobulin products could be prepared to ensure the amount of neutralizing antibodies and facilitate stable supply, storage, and administration. The effect and safety of CP and hyperimmune immunoglobulin are still controversial and need further investigations .
The sero-kinetic evaluation of the two COVID-19 patients with pneumonia suggests several implications about the serologic test method we used. First, the sensitivity of IgG and total ELISA were quite sensitive, but the COI values and OD ratios peaked at the very early phase, which might be due to saturation of signals. Meanwhile, FIA IgM showed lower COI values compared to IFA IgG, but a linearity during the course of illness was observed in Patient A.
This suggests that sensitive immune assay methods could be saturated early, and titration would be required for the evaluation of sero-kinetics of COVID-19 patients . Second, based on the present study’s findings and previous reports, IgM might play a limited role in terms of serologic diagnosis. IgG and IgM converted to be positive at a similar time point in Patient A, though sera between day 5 and 9 were not available.
In the serial evaluation of Patient B, the FIA IgM COI values were lower than those of IgG and showed a waning trend before the end of viral shedding. In previous reports, IgM antibodies were detected two days earlier than IgG antibodies, and IgM antibodies waned earlier than IgG antibodies [23,34].
The window period that IgM appears earlier than IgG would be very narrow, and may not even exist in mild patients in whom IgM response weakly occurs . In addition, IgM antibodies in convalescent sera of mild patients were limitedly detected in the present analysis, which is compatible with other published reports .
Third, on the other hand, severe patients may show a suppressed antibody response. Compared to Patient A, who recovered from pneumonia without mechanical ventilator support, Patient B showed lower IgM COI values and MN titer. Although each sero-kinetics study did not include a sufficient number of patients according to detailed severities, decreased antibody response in critical COVID-19 patients was also noticed in previous reports .
We note that our study has several limitations.
First, patient numbers of asymptomatic/anosmia and pneumonia groups were relatively limited according to the disease severity classification of the present study. However, we noticed that antibody production occurred even in asymptomatic patients, while high titer production was limited in mild groups.
Second, there could be a recall bias regarding the symptoms of the patient since they were recorded after recovery from the disease. Nevertheless, initially negligible symptoms were additionally identified by telephone interview. About a half of patients who were initially classified as asymptomatic were not truly asymptomatic and recalled mild symptoms in the telephone interview, especially anosmia.
Lastly, IgM antibodies were limitedly measured in the present investigation. Since we only use a single test method for IgM and did not evaluate serial serum specimens from mild patients, it is not clear whether IgM antibodies were scarcely formulated in mild disease, diminished early, or sensitivity of FIA for IgM antibodies was low.
However, as discussed above, it is more likely that IgM antibody production is limited in mild COIVD-19 patients considering other reported studies. This could be further evaluated by a sero-kinetics study for asymptomatic and mild COVID-19 patients.
reference link : https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7408950/