Development and clinical application of a rapid SARS‐CoV‐2 antibody test strip: A multi‐center assessment across China

Abstract Background The ongoing coronavirus disease 19 (COVID‐19) is posing a threat to the public health globally. Serological test for SARS‐CoV‐2 antibody can improve early diagnosis of COVID‐19 and serves as a valuable supplement to RNA detection. Method A SARS‐CoV‐2 IgG/IgM combined antibody test strip based on colloidal gold immunochromatography assay was developed, with both spike protein and nucleocapsid protein of SARS‐CoV‐2 antigen used for antibody detection. From 3 medical institutions across China, serum or plasma of 170 patients with confirmed COVID‐19 diagnosis and 300 normal controls were collected and tested with the strip. Sensitivity, specificity, kappa coefficient, receiver operating characteristic (ROC) curve, and area under the curve (AUC) were analyzed. Positive rates in different medical centers, age group, gender, and different disease course were compared. Results 158 out 170 samples from confirmed COVID‐19 patients had positive results from the test, and 296 out of 300 samples from normal controls had negative results. The kit was 92.9% sensitive and 98.7% specific. The positive rate was 77.3% during the first week after disease onset, but reached 100% since day 9. AUC and kappa coefficient were 0.958 and 0.926, respectively, which showed the consistency of the test results with the standard diagnosis. Age or gender caused little variations in the kit sensitivity. Conclusion The rapid, easy‐to‐use SARS‐CoV‐2 IgG/IgM combined antibody test kit has a superior performance, which can help with accurate diagnosis and thus timely treatment and isolation of COVID‐19 patients, that contributes to the better control of the global pandemic.


| INTRODUC TI ON
Coronavirus disease  caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is rapidly spreading on a global scale and poses a grave threat to the public health. As a highly contagious disease, it caused over 26,000,000 confirmed cases of infection and over 870,000 deaths as of early September 2020, which are still on a constant rise. 1 Though the mortality of COVID-19 is around only 3.3%, the various ways of transmission including droplets, aerosol and fecal transmissions, as well as the large population base of patients, make it urgent to develop approaches to early identifying and controlling the disease. [2][3][4] The diagnosis of COVID-19 has been largely dependent on nucleic acid RT-PCR with patient throat swab. However, the sensitivity of this test is insufficient, due to the low viral loads in upper respiratory tract of patients, failure to correctly collect high-quality swab specimen, and varied viral loads in different stage of infection. Liang was proposed to test the specific antibody in the patient blood as a supplement to nucleic acid RT-PCR. 6 IgM antibody can be detected 3 days after the illness onset and lasts until the patient recovers; IgG antibody appears on day 3-7 and exists constantly even in the convalescence. [7][8][9] Another report which compared different approaches of detections showed that the total sensitivity of RT-PCR was 67.1%, while the antibody tests had a sensitivity of 93.1%. Surprisingly, it was up to 99.4% sensitive when combining antibody test with nucleic acid RT-PCR results. 6 With the emphasis on the antibody test, the WHO interim guidelines updated on 19th of March approved the need for antibody testing to support COVID-19 diagnosis. 10 Accordingly, the 7 edition of guidelines for diagnosis and treatment of COVID-19 published by National Health Committee of China has included serological test of specific antibody of SARS-CoV-2 in the diagnosis criteria, which suggests that in case where a suspected patient gets negative result from PCR, a positive serological antibody test serves as an evidence to confirm the diagnosis. 11 Further, COVID-19 vaccine is under rapid development and regarded as a priority for ending the pandemic, which again indicates the importance of antibody detection, as it is needed for the evaluation of response to the vaccine candidates. 12,13 There has been quite a few detection kits for serological test of SARS-CoV-2 antibody. 14-17 However, using a single protein as the probe limited the detection sensitivity. Spike (S) protein is located on the surface and able to bind with host cell membranes.
Nucleocapsid (N) protein is an abundant protein which structurally binds to RNA. [18][19][20] Both S and N proteins are highly immunogenic and antibodies against these two can be detected in patients with COVID-19, but most serological test kits target on either S or N protein alone. 15 In addition to this, IgG or IgM detection does not work as good as combined IgG/IgM detection. Here we developed a SARS-CoV-2 antibody test strip based on colloidal gold and immunochromatography assay, with using both spike protein and nucleocapsid protein of SARS-CoV-2 as antigen, which detects both IgG and IgM antibody instead of only one of them. A multi-center assessment across China was carried out on 470 individuals, consisting of 170 confirmed COVID-19 patients (by RT-PCR and clinical manifestations) and 300 negative controls. With this rapid and sensitive SARS-CoV-2 IgG/IgM combined antibody test kit, we hope to further improve the accuracy of diagnosis for COVID-19 patients, thus contributing to better control of the worldwide pandemic.

| Materials for the test strip
The recombinant antigens of SARS-CoV-2 were commercially purchased-S protein (S1-RBD) from GenScript Inc (Cat#: T80302) and N protein from Fapon Biotech Inc (Cat#: ncov-PS-Ag6). Both S protein and N protein were used as the antigen in the kit. Detail

| Preparation of colloidal gold and conjugation of SARS-CoV-2 antigen
Purified water was brought to boil in a round-bottom flask, when reductive agent trisodium citrate dihydrate solution was added and kept boiling for 1 minute. Tetrachloroauric acid solution was then added and kept boiling for 10 minutes. The solution was cooled down to room temperature, and purified water was added to make 800 milliliter of total volume. The final colloidal gold solution was checked, and those with good permeability under the light were deemed qualified. Collect the solution in a clean flask until use.
Potassium carbonate was added to colloidal gold solution and stirred for 20 minutes. While stirring, N protein of SARS-CoV-2 antigen was added to make final concentration of 10μg/mL and kept stirring for 45 minutes. Centrifuge at 12 000 rpm for 50 minutes and remove the supernatant. Collect the conjugation solution, centrifuge again, and remove the supernatant, and the final colloidal gold conjugated with SARS-CoV-2 N protein was obtained. Apply the same protocol to make colloidal gold conjugated with SARS-CoV-2 S protein and with goat anti-chicken IgY polyclonal antibody. Mix all the conjugates and spray to the conjugate pad.

| Test principal of SARS-CoV-2 antibody test strip
The SARS-CoV-2 antibody test strip was based on colloidal gold im-

| Sample testing
A dropper was inserted the into the blood collection tube and absorbed the sample automatically until it reached to 10μL. The sample was added to the sampling area on the strip, and 2 drops of sample diluent were added ( Figure S1A). 15-20 minutes later, the results were read. The interpretation of the results is shown in Figure S1B. where P < .05 was considered statistically significant.

| Production of SARS-CoV-2 antibody test strip and result display of clinical samples
The product of test strip for specific SARS-CoV-2 antibody based on colloidal gold and immunochromatography assay was completed  Figure 1A,B.

| Demographic information of participants for clinical assessment
The clinical information of these participants is summarized in

| Evaluation of SARS-CoV-2 antibody kit performance
Of the patients with detailed information, a timeline graph was depicted to illustrate the illness onset, hospital admission, serological test for antibody, and discharge of each patient, along with their clinical classification (Figure 2). In all the 28 patients, most were positive for antibody test except for 2 patients who received the test within a week after symptom onset.
Among the total 170 samples with confirmed cases from all medical centers, 158 had positive result from the test, generating a sensitivity of 92.9%. Among the 300 samples from normal controls, 296 had a negative result, with a specificity of 98.7% (Table 2).

Province, Chest Hospital of Shandong Province and Infectious
Diseases Hospital of Xuzhou City, the sensitivity were 97.6%, 87.3%, and 95.4%, respectively, and the specificity was 96.9%, 100%, and 99.1%, respectively. Further, to evaluate the consistency of the test result and the standard diagnosis, Youden index, positive predictive value, negative predictive value, and kappa coefficient were calculated for all samples and each center (Table 2). ROC curve for all participants was obtained and shown in Figure S2A, and ROC curves for each medical center were shown in Figure S2B-D. AUC and kappa coefficient were summarized as well ( Table 2). These data verified  Despite analysis in total participants, the dynamic changes of positive rates were depicted with the information on disease course of each patient. The positive rate increased during the first week after illness onset, reached 100% on day 9 and remained unchanged ( Figure 3). Further, the kit sensitivity in different medical centers, gender, age, and disease course was compared, and little variation with kit sensitivity was found ( Figure 4A-C), except for that it was lower for samples collected within 1-7 days after symptom onset ( Figure 4D).

| D ISCUSS I ON
In this study, we developed a rapid, convenient, and easy-to-interpret kit for SARS-CoV-2 specific IgG/IgM antibody. With these advantages, the test kit can be applied for early diagnosis of SARS-CoV-2 infection among the suspected patients, medical staff, and population with possible exposure history, which ensures they get the treatment and isolation in time. Despite some argument that the antibody kit had limited efficacy in community screening, it is officially recommended by WHO as a diagnostic tool for COVID-19. 10,23 Another application of antibody test is the treatment-related evaluations. COVID-19 vaccine is the most attractive approach to eradicate the virus globally, and it is moving toward the preclinical and clinical trials. 24,25 As active immunization generated, the serological antibody can be detected, which indicates the immunity against SARS-CoV-2 infection. Additionally, monoclonal antibody therapy was raised as a promising intervention for COVID-19, and the transfusion of convalescent plasma was proved to be beneficial in critically ill patients. [26][27][28][29] This also suggested the need for antibody test as the convalescent plasma should be assessed before the transfusion was carried out.
Further, the serological test is potentially useful to identify asymptomatic infection. There is a largely unknown but presumably high proportion of asymptomatic carriers of SARS-CoV-2. A recent report showed 12% of asymptomatic (SARS-CoV-2 detected but symptoms never develop) and 30% of presymptomatic cases (SARS-CoV-2 detected before symptom onset) in travelers and returning residents to Brunei. 30  In conclusion, we developed a rapid and easy-to-use SARS- CoV-2 antibody test kit which is highly sensitive and specific in the diagnosis for COVID-19. With the widespread use of the kit, earlier treatment and timely isolation can be achieved, that improves the outcome and controls the extensive transmission of the global pandemic.

ACK N OWLED G M ENTS
We appreciate the medical staff from the Center for Disease Control

CO N FLI C T S O F I NTE R E S T
The authors declare no conflicts of interest.

F I G U R E 3
Dynamic changes of positive rates of the test kit. The positive rate increased within the first week after illness onset, and reached 100% on day 9 F I G U R E 4 Sensitivity comparison among results from different patient groups. Column chart that compares sensitivity in different medical centers (A), gender (B), age group (C) and disease course (D). Data were shown as mean with 95% confidential interval (CI). P < .05 was considered as statistically significant. CDCJS: The Center for Disease Control and Prevention of Jiangsu Province; CHSD: Chest Hospital of Shandong Province; IDHXZ: Infectious Diseases Hospital of Xuzhou City