key: cord-1024444-dplcwwps
authors: Başaran, Seniha; Şimşek-Yavuz, Serap; Meşe, Sevim; Çağatay, Atahan; Medetalibeyoğlu, Alpay; Öncül, Oral; Özsüt, Halit; Ağaçfidan, Ali; Gül, Ahmet; Eraksoy, Haluk
title: The effect of tocilizumab, anakinra, and prednisolone on antibody response to SARS-CoV-2 in patients with COVID-19: A prospective cohort study with multivariate analysis of factors affecting the antibody response
date: 2021-03-15
journal: Int J Infect Dis
DOI: 10.1016/j.ijid.2021.03.031
sha: 3961acf145eff79d984b4b090323b582c2f2f117
doc_id: 1024444
cord_uid: dplcwwps

OBJECTIVES: Disease severity, previous medications, immunosuppresive agents could affect the antibody response against SARS-CoV-2. We aimed to analyze variables affecting the humoral response to SARS-CoV-2. METHODS: In this prospective cohort study, we included adult patients who recovered from COVID-19 and were admitted to COVID-19 follow-up unit. We defined 8 patient groups in accordance with the results of thorax CT, SARS-CoV-2 PCR test, and tocilizumab or anakinra use during active disease. Anti-S IgG antibodies were determined by ELISA in serum samples. Anti-S positive and negative cases were compared. RESULTS: A total of 518 patients were included in the study. SARS-CoV-2 IgG antibodies were positive in 82.8% of patients. SARS-CoV-2 PCR positivity, extent of lung involvement on CT, and time to antibody testing were independently associated with antibody positivity. Tocilizumab, anakinra or prednisolone use was not a factor affecting the antibody response. The rate of antibody response and sample/CO values among antibody positive patients showed a linear relationship with the extent of lung involvement on CT. CONCLUSIONS: The use of tocilizumab, anakinra, and prednisolone for COVID-19 did not affect the antibody response against SARS-CoV-2. The main driver of antibody response among patients with COVID-19 was the extent of pulmonary involvement on CT.

After the publication of some observational studies showing beneficial effects of anti-cytokine treatments for COVID-19, biologic agents blocking the activity of IL-6 such as tocilizumab and activity of IL-1, such as anakinra, have been widely used to treat COVID patients who developed findings of cytokine storm syndrome (Toniati et al., 2020) . In addition, glucocorticoid treatment has become a standard of care for severe COVID-19 patients after demonstration of the beneficial effect of dexamethasone on mortality (RECOVERY Collaborative Group, 2020). All of these immunosupresive treatments could affect the antibody response against SARS-CoV-2 (Maeda et al., 2010 , Roll et al., 2011 , Dinarello, 2009 ).

In this study, we analyzed all potential factors which may affect the antibody response to SARS-CoV-2 in a cohort of patients with different levels of disease severity and SARS-CoV-2 PCR results, and specifically aimed to investigate whether or not this response was affected by tocilizumab, anakinra or by prednisolone treatments.

In this prospective cohort study, we included all consecutive adult patients who were diagnosed with probable or confirmed COVID-19 in our hospital, recovered from the disease, admitted to our COVID-19 Follow-up Unit between June, and September 2020 and fulfilled the below specific group criteria. The definitions of the World Health Organisation for COVID-19 case were used: A patient who meets the appropriate clinical criteria (acute onset of fever and cough; or acute onset of any three or more of the following signs or symptoms of fever, cough, general weakness/fatigue, headache, myalgia, sore throat, coryza, dyspnoea, anorexia/nausea/vomiting, diarrhoea, altered mental status) and residing or travel to an area with community transmission anytime within the 14 days prior to symptom onset and either with chest imaging showing findings were suggestive of COVID-19 disease and/or with a positive SARS-CoV-2 PCR test result (WHO, 2020). To include all patients with different disease severity and microbiologic result status, we defined 8 patient groups in accordance with the results of thorax computed tomography (CT), SARS-CoV-2 PCR test and use of tocilizumab or/and anakinra during the active phase of COVID-19 disease before the enrollment into the study: Group 1: Patients treated with at least 1 dose of tocilizumab and/or anakinra for COVID-19; Group 2: Patients with a positive SARS-CoV-2 PCR test, and had severe involvement on thorax CT; Group 3: Patients J o u r n a l P r e -p r o o f Physical examination was performed, and blood samples were drawn from patients after the consents were obtained. The serum samples were stored at -20 0 C until antibody testing. The epidemiological, clinical features, and admission laboratory tests and thorax CT results of patients were retrospectively obtained from the online database of the hospital.

Statistical analysis was performed using the Statistical Package for the Social Sciences (SPSS) 21 .0 (SPSS Inc., Chicago, IL, USA). Univariate analysis was done using Chi-square, Fisher's exact, Student's t, and Mann-Whitney U tests, where appropriate. For the multivariate logistic regression analysis, significant (with a p value of < 0.05) factors identified with univariate analysis were included to the model to determine the independent predictors of SARS-CoV-2 spike IgG antibody seroconversion. Colinear variables were excluded, collinearity between predictors was tested by correlation matrix. All factors included into the multivariate logistic regression analysis were shown in Table 1 . Hosmer-Lemeshow goodness of fit statistics were used to assess the model fit. Kruskal-Wallis tests were conducted to compare the non-normally distributed parameters and ordinal variables between the groups, and the level of pulmonary involvement on thorax CT.

A total of 518 patients (63.3% hospitalized, 36.7% outpatient) were included in the study. The IgG antibodies were positive in 82.8% of patients, the rate was 90% and 72% among PCR positive and negative cases, respectively; and 94% of PCR positive patients with no antibody response had either no (52%) or minor (42%) involvement on thorax CT. The demographic, and clinical features of patients with and without antibody response are shown in Table 1 .

In univariate analysis, solid organ transplantation (p=0.003), the use of immunosuppresive drugs before the COVID-19 diagnosis (p=0.01); presence of fever, low WBC, neutrophil and lymphocyte counts, high CRP and ferritin levels on admission; need for oxygen support and hospitalisation, severity of involvement on thorax CT, nasopharengeal SARS-CoV-2 PCR positivity, time from hospital admission to antibody testing, tocilizumab or anakinra or prednisolone use for COVID-19 were found to be related with antibody positivity (Table 1 ). In multivariate analysis, SARS-CoV-2 PCR positivity and extent of involvement on thorax CT and time from hospital admission to antibody testing were found independently associated with antibody positivity ( Table 2 ).

The analysis of only the confirmed COVID-19 cases (with a positive result of SARS-CoV-2 PCR test) showed that older age (p=0.047), hospitalisation (p<0.001), presence of fever (p=0.014), and pneumonia (p=0.001), higher serum CRP (p<0.001), and ferritin levels (p<0.001), lower blood WBC (p=0.001), and PNL counts (p=0.032), moderate/severe lung involvement on thorax CT (p=0.001), the need for oxygen support (p=0.002), use of favipiravir (p=0.001) and tocilizumab and/or anakinra (p=0.012) were found as the factors associated with antibody response. Although serum D-dimer level was found higher, and blood lymphocyte count was found lower among patients with a positive SARS-CoV-2 spike IgG, the differences did not J o u r n a l P r e -p r o o f reach to a statistical significance (p=0.064, and 0.07, respectively) ( Table 1) . Moderate/severe lung involvement on thorax CT was found as an independent factor of SARS-CoV-2 spike antibody positivity also among confirmed cases (OR 10.95, 95% CI 1.20-99.81, p=0.034) ( Table   2 ).

Both the rate of antibody response and sample/Cut-off (S/CO) values among antibody positive patients showed a linear relationship between the extent of lung involvement on thorax CT ( Figure 1 ) and both were significantly different between the groups (p <0.05). S/CO level was found to be 2.21, 1.69 and 1.16 times higher in patients with severe, mild and moderate involvement compared to no lung involvement, respectively. Groups including severe COVID-19 cases (Group 1-3) showed significantly higher titer responses compared with the groups including milder cases (Group 5-8) (p <0.001). The antibody concentrations among patients with a positive antibody result (429 patients) were significantly higher in patients who were treated with either tocilizumab, or anakinra or prednisolone (p <0.05). However, S/CO levels of those patients were found to be similar with the S/CO levels of patients with severe disesase (p >0.05) ( Table 3 ).

We evaluated the factors affecting antibody response against SARS-CoV-2 in a cohort of COVID-19 patients with different level of severity and PCR test results, and with and without tocilizumab and/or anakinra and/or prednisolone treatments in this study. We found that antibody response was independently associated with the level of lung involvement on thorax CT. 95%, 90%, 69% of patients with severe, moderate and mild lung involvement, respectively were positive for SARS-CoV-2 antibody. S/CO level was found to be 2.21, 1.69 and 1.16 times higher J o u r n a l P r e -p r o o f in patients with severe, mild and moderate CT involvement than in patients with no involvement, respectively. When we limited the analysis to only confirmed cases, we also found the moderate/severe involvement of lung on the thorax CT as an independent factor affecting the antibody seroconversion. The extent of involvement on thorax CT was shown to be directly However, we found no other study showing a direct relationship between the extent of lung involvement on thorax CT and the rate and level of antibody response. Therefore, based on the findings of the current study, antibody responses and level of it could be predicted from the level of lung involvement on thorax CT, even in patients without a positive PCR test.

We also found that although antibody response rate was higher in patients who were treated with either tocilizumab, anakinra, or prednisolone compared with the patients who were not administered these drugs in univariate analysis. However, the use of those anticytokines and prednisolone for the treatment of COVID-19 were not found as independent factors affecting the antibody response rate. There could be so many reasons why we failed to obtain less SARS-CoV-2 antibody seroconversion in those COVID-19 patients who were treated with drugs affecting the immune system. First of all, it is well known that potential adverse effects of immunosupressive or immunomodulatory agents is related to both the average dose and the cumulative duration of use (Slade and Hepburn, 1983 ). In the case of COVID-19, those drugs were used no more than 1 or maximum of 2 weeks and as a result, this could be one of the J o u r n a l P r e -p r o o f reasons why we found no adverse effect of those drugs on antibody response to SARS-CoV-2.

Our finding of lower antibody response among COVID-19 patients who were on long term immunosuppressive drugs also supports that idea. IL-6 works as a B cell differentiation factor, which induces activated B cells to produce immunoglobulin (Muraguchi et al., 1988) . T cell-dependent antibody response against virus infection is impaired in IL-6 deficient mice (Kopf et al., 1994) . But we found that tocilizumab, an IL-6 receptor blocker did not affect the antibody response in patients with COVID-19. It was also reported in another study that tocilizumab did not impair the viral specific antibody The key to success of either anti-inflammatory or immunostimulatory type of immunomodulation is to identify the patients who may benefit from a particular intervention. Considering the complexity of the host response during COVID-19 and the diversity of pathophysiologic mechanisms at play, it is unlikely that the current"one-target"

and "one-size-fits-all" strategy will control all part of the dysregulated immune system (Van der Poll and Wiersinga, 2020). Another explanation may be that tocilizumab and/or anakinra are not targeting B cell specifically, antibody production may not depend simply on IL-1, or IL-6. IL-6 was not required for influenza virus specific antibody responses by tonsillar mononuclear cells in an in vitro study (Costelloe et al., 1993) . Lastly, inflammatory environment in severe patients with cytokine storm results in exhaustion of lymphocytes and lymphopenia, and improvement of inflammatory status may be associated with better immune functions against SARS-CoV-2 both in humoral, and cellular immunity (Riva et al., 2020).

Although glucocorticoids cause no significant acute changes on B cells (Olnes et al., 2016) , the number of circulating B lymphocytes and levels of IgG may be reduced with shortterm glucocorticoid administration (Slade and Hepburn, 1983 ). In our study, prednisolone used for COVID-19 treatment was not found to be associated with the antibody response. In a study, hospitalized COVID-19 patients received corticosteroids and/or tocilizumab, no detrimental effect was found on antibody responses. Although differences in antibody response even favored patients receiving corticosteroids, such differences vanished after adjustment (Masiá et al., 2020b) . It was same in our study, patients treated with either tocilizumab, anakinra, or J o u r n a l P r e -p r o o f prednisolone had higher antibody response than patients who did not take these medications, but none were found as an independent factor for antibody response in multivariate analysis. As a result, we may conclude that, tocilizumab, anakinra and prednisolone, even in combination, do not negatively impact the humoral immune response against SARS-CoV-2.

We found that seroconversion in COVID-19 cases were related to the duration between hospital admission, and antibody testing (OR 1.02; 95% CI 1.01-1.03; p <0.001). It was shown that SARS-CoV-2 IgG antibodies develop over a period of 7-50 days from the symptom onset, with a median of 24 days (Wajnberg et al., 2020a). Although the mean time from hospital admission to antibody testing was 82 days in our study, antibody testing was done before 42 days in 25% of our patients, and as a result some of them could not produce adequate antibody in a given time. As 76% of our patients with a negative antibody response have mild disease either with no involvement or mild involvement on thorax CT, another explanation of time dependence of seroconversion could be the longer duration of antibody response among asymptomatic and mild cases. It was also shown that seroconversion in mild COVID-19 cases might take longer time to mount (Wajnberg et al., 2020a (Wajnberg et al., , 2020b . Another explanation for this could be due to a mild and transient response, antibodies become negative in a short time among mild or asymptomatic cases (Poland et al., 2020) . In the convalescent phase, IgG titres in symptomatic individuals remained significantly higher than those in asymptomatic individuals (Adams et al., 2020) . Notably, in one study IgG titres declined during the convalescent phase in both symptomatic and asymptomatic individuals, with 13% and 40% of symptomatic and asymptomatic individuals becoming IgG seronegative within 2-3 months following infection (Long et al., 2020) . Finally, some of our patients might not have COVID-19 at all as the rate of antibody positivity of our patients was found as 72% without a PCR positivity. However, 31 out of 89 antibody negative patients were confirmed COVID-19 cases with a positive PCR test. Our analysis of the only confirmed COVID-19 cases, showed that the time from hospital admission to antibody testing was not found as a factor affecting the antibody response. This could be related to the very early sampling and false negativity of PCR testing among PCR negative cases, which could be the result of delayed antibody response among those cases. As we did not plan to assess the kinetics of the antibody response we cannot make a final comment on this issue.

Our study has some limitations. Although we included the patients prospectively during convalescent phase, the clinical and laboratory data of active COVID-19 were obtained from hospital database retrospectively, as a result we could not get some of the data including the duration of symptoms, total duration of prednisolone use, and the time from the symptoms to thorax CT evaluation. As SARS-CoV-2 RNA PCR was negative in 40% of our patients, some of them might not have COVID-19 at all. But we included also the thorax CT results for our analysis, and 72% of our patients with a negative PCR test had a positive antibody test result, which shows that selection of patient is appropriate.

Our study has some strengths. We analyzed factors affecting antibody response in a high number of patients with different disease severity, and PCR result status. For the first time, we showed a direct association between antibody response rate and S/CO ratio and extent of lung involvement on CT. We also showed that anakinra treatment does not affect the antibody response in patients with severe COVID-19.

J o u r n a l P r e -p r o o f antibody response, their use during the course of COVID-19 was not found to be associated with a negative effect on antibody response.

The main driver of the antibody response to SARS-CoV-2 and level of it among patients with COVID-19 was found to be the extent of lung involvement on thorax CT, which is correlated well with the severity of infection, and as a result we could predict the antibody response to SARS-CoV-2 and levels of it by thorax CT.

SB, SŞY and SM conceived of and designed the study. SB and SŞY analysed the results, and wrote the manuscript. All authors reviewed and approved the final version for submission.

University (Project number: 36884).

No conflict of interest was declared by the authors. 

Antibody testing for COVID-19: a report from the National COVID Scientific Advisory Panel

Quick COVID-19 healers sustain anti-SARS-CoV-2 antibody production

Interleukin 6 is not required for antigen-specific antibody responses by human B cells

Immunological and inflammatory functions of the Interleukin-1 family

Early prediction of disease progression in COVID-19 pneumonia patients with chest CT and clinical characteristics

Personalized therapy approach for hospitalized patients with COVID-19

Relationship between chest CT manifestations and immune response in COVID-19 patients

Chest CT in COVID-19 pneumonia: A review of current knowledge

Impaired immune and acute-phase responses in interleukin-6-deficient mice

Clinical and immunological assessment of asymptomatic SARS-CoV-2 infections

IL-6 increases B-cell IgG production in a feed-forward proinflammatory mechanism to skew hematopoiesis and elevate myeloid production

Impact of interleukin-6 blockade with tocilizumab on SARS-CoV-2 viral kinetics and antibody responses in patients with COVID-19: a prospective cohort study

Lack of detrimental effect of corticosteroids on antibody responses to SARS-CoV-2 and viral clearance in patients hospitalized with COVID-19

Impact of tocilizumab therapy on antibody response to influenza vaccine in patients with rheumatoid arthritis

The essential role of B cell stimulatory factor 2 (BSF-2/IL-6) for the terminal differentiation of B cells

Effects of systemically administered hydrocortisone on the human immunome

SARS-CoV-2 immunity: review and applications to phase 3 vaccine candidates

A multicentre, randomised, double-blind, placebo-controlled trial with the interleukin-1 receptor antagonist anakinra in patients with systemic-onset juvenile idiopathic arthritis (ANAJIS trial)

Dexamethasone in hospitalized patients with Covid-19 -preliminary report

COVID-19: room for treating T cell exhaustion?

Invivoeffects of the anti-interleukin-6 receptor inhibito rtocilizumab on the B cell compartment

Antigen-specific adaptive immunity to SARS-CoV-2 in acute COVID-19 and associations with age and disease severity

Prednisone-induced alterations of circulating human lymphocyte subsets

Efficacy of tocilizumab in patients hospitalized with Covid-19

Tocilizumab for the treatment of severe COVID-19 pneumonia with hyperinflammatory syndrome and acute respiratory failure: A single center study of 100 patients in

Sepsis and septic shock

Humoral response and PCR positivity in patients with COVID-19 in the New York City region, USA: an observational study

Robust neutralizing antibodies to SARS-CoV-2 infection persist for months

World Health Organization (WHO)

The authors thank Yunus Çatma for his help in data collection.J o u r n a l P r e -p r o o f