key: cord-0855162-fsj34b39
authors: Patel, M.; Lakhotia, Y.; Shah, S.; Suthar, N.; Kola, S.; Rangarajan, R.
title: Clinical validation of a point-of-care antibody test for COVID-19
date: 2020-12-18
journal: nan
DOI: 10.1101/2020.12.16.20248303
sha: d36fc54e2088eff23318313c970c1d02d6efb7cf
doc_id: 855162
cord_uid: fsj34b39

The objective of this study was to evaluate the performance of a lateral flow antibody test for COVID-19, approved for use in India. Although many point-of-care antibody tests are available globally, they have been subjected to limited clinical validation. This has led to suboptimal outcomes in the field, where antibody tests play a significant role in tracking the immunity of individuals and communities. In this study an antibody test, ImmunoQuick that recognizes antibodies to the Nucleocapsid and Spike proteins of SARS CoV-2 was tested in 100 symptomatic patients with a positive or negative diagnosis of COVID-19, based on RT-PCR results. The overall sensitivity of the test was found to be 86.1% (95% CI: 76.4% to 92.8%) and specificity 100% (95% confidence interval: 73.5% to 100%). The sensitivity reached a peak of 95.7% with samples taken 17 days after the onset of symptoms. Overall, the sensitivity and specificity of the test are sufficient for assessing seroprevalence.

The COVID-19 pandemic is expected to persist beyond 2020, despite global efforts to contain the 27 infection. Tracking the extent of spread is one of the key ways in which policy decisions regarding the 28 opening of businesses and educational institutions can be made. This involves measuring the exposure of 29 individuals to the virus, assessing protection from reinfection and monitoring immunity at the population 30 level. Point-of-care (POC) antibody tests are well suited for this purpose; their cost and ease of use allow 31 for repeated testing in clinical and non-clinical settings. Several POC antibody tests based on lateral flow 32 immunochromatography have been approved for use globally (1) . However, not all tests are backed by 33 robust validation data. Hence, there have been questions about their sensitivity and specificity when 34 deployed in the field (2, 3, 4, 5) . In India, one of the earliest tests approved was from ImmunoScience India 35

Private Limited. We undertook a clinical study to confirm the performance of ImmunoScience's lateral 36 flow immunoassay, ImmunoQuick, that detects IgG and IgM antibodies against the Nucleocapsid and 37

Spike proteins of the SARS CoV-2 virus, from capillary blood (6). The study was conducted in a cohort of 38 individuals with influenza-like symptoms, who were hospitalized at the Sardar Vallabhbhai Patel Institute 39 of Medical Science and Research, Ahmedabad in the state of Gujarat. The objective of the study was to 40 measure the sensitivity and specificity of the assay in individuals who had a confirmed diagnosis of being 41 positive or negative for COVID-19. The data showed an overall sensitivity based on the detection of IgG 42

and IgM antibodies as 86.1% (95% CI: 76.4% to 92.8%) and specificity 100% (95% confidence interval: 43 73.5% to 100%). These performance characteristics are sufficient for tracking and monitoring the 44 immunity in individuals. 45

Registry of India with the registration number, CTRI/2020/07/026469. The study was approved by the 48 All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. were discharged while those with a positive result were kept hospitalized until 10 days past the onset of 54 symptoms or until the resolution of symptoms. 55 A total of 100 subjects were enrolled in the study, 80 positive and 20 negative for SARS CoV-2 by RT-PCR. 56 All subjects were above the age of 18. Those who had tested positive were at least 10 days past the onset 57 of symptoms at the time of enrollment. Those who were negative by RT-PCR were mostly between day 3 58 to 9 from the onset of symptoms (14/20); 4 were between 1 to 2 days and 2 were between 16 to 17 days 59 from the onset of symptoms. Individuals who were critically ill or had a history of immunosuppression or 60 were participating in any other clinical trial were excluded. 61 62 Study design 63

Informed consent was taken from eligible individuals in writing. Basic demographic information, medical 64 history and RT-PCR results were recorded. Then the lateral flow antibody test was performed by lancing 65 the index finger and applying two drops of blood in the sample window of the test cassette, followed by 66 two drops of buffer, as per the manufacturer's instructions. Results were read between 15 and 20 minutes 67 after the application of blood. If individuals had a negative RT-PCR result but were positive by the lateral 68 flow assay, venous samples were taken and frozen. These samples were then processed using an Enzyme-69 linked immunosorbent assay (ELISA) kit for total antibodies (COVIDscreen Plus) from Trivitron Healthcare 70

Pvt. Ltd. A schematic view of participant flow is shown in Figure 1 . 71 All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this this version posted December 18, 2020. of sensitivity and specificity were done as per the following formulae: 76 All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Assay sensitivity 87 All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this this version posted December 18, 2020. ; https://doi.org/10.1101/2020.12.16.20248303 doi: medRxiv preprint were positive for either IgG or IgM antibodies by the Immunoscience lateral flow assay ( Table 1 ). The 89 sensitivity was calculated to be 86.1%, with a 95% confidence interval (CI) of (76.4% to 92.8%). 90 An analysis of sensitivity based on the number of days from onset of symptoms showed time dependent 93 changes. When considering IgM positivity alone, sensitivity improved from 65.5% to 84.6% and dropped 94 to 69.6% for samples taken at 10-13 days, 14-17 days, and 18-40 days respectively from the onset of 95 symptoms. This trend is consistent with reported data on the rise and fall of IgM levels in patients, 96

following a COVID-19 infection (7). When considering IgG positivity alone, sensitivity improved from 50.6% 97 to 80.8% to 91.3% for samples taken at 10-13 days, 14-17 days, and 18-40 days respectively from the 98 onset of symptoms. Finally, when considering overall sensitivity (IgM or IgG positivity), sensitivity 99 improved from 75.9% to 82.3% to 95.7% for samples taken from patients at 10-13 days, 14-17 days, and 100 18-40 days respectively from the onset of symptoms (Figure 3 ). (A single data point from a patient at day 101 73 and negative for IgM and IgG antibodies was excluded). These data suggest that the sensitivity of the 102 assay depends on the number of days from onset of symptoms. Highest sensitivity is achieved 17 days 103 after the onset of symptoms. 104

In this study, we noted 11 false negative results. 9 out of 11 samples were taken between 10 and 15 days 105 from the onset of symptoms, one at 19 and one at 73 days. It is possible that the antibody titer in these 106 individuals was below the limit of detection of the lateral flow assay. 107 All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this this version posted December 18, 2020. ; https://doi.org/10.1101/2020.12.16.20248303 doi: medRxiv preprint Figure 3 : Test sensitivity with respect to days from onset of symptoms 111

In this study, individuals were tested only once. Thus, a time course of antibody responses could not be 113 derived. However, we analyzed the relationship between the number of days from the onset of symptoms 114 and the type of antibody positivity (Table 3 ). The number of RT-PCR positive samples that were positive 115 for IgM alone was 9, positive for IgG alone were 11 and positive for IgM and IgG were 48. The median 116 All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this this version posted December 18, 2020. ; https://doi.org/10.1101/2020.12.16.20248303 doi: medRxiv preprint days, with overlapping ranges for the three types of immune profiles. Most individuals were positive for 118

IgM and IgG antibodies, which is consistent with the time course of antibody responses reported in the 119 literature (7). 120 Table 3 Of the 20 samples that were negative by RT-PCR, only 12 were negative by the lateral flow assay as well 124 (Table 4 ). 8 samples, taken between days 1 to 9 from onset of symptoms, showed distinct IgM positive 125 signals ( Figure 4) . As per the study protocol, venous blood from these individuals was analyzed by ELISA. 126

All 8 were positive for antibodies (data not shown), taking the specificity to 100%, with a 95% confidence 127 interval of 73.5% to 100%. 128 preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this this version posted December 18, 2020. ; https://doi.org/10.1101/2020.12.16.20248303 doi: medRxiv preprint In this study we evaluated ImmunoScience's lateral flow immunochromatography assay for the detection 138 of IgM and IgG antibodies specific to SARS CoV-2. We found the specificity to be 100% (95% CI: 73.5% to 139 100%), while the sensitivity was 86.1% (95% CI: 76.4% to 92.8%). Peak sensitivity of 95.7% was achieved 140 after day 17. These results suggest that the test adequately serves the purpose of identifying those who 141 have been exposed to the virus. 142 majority of the samples were positive for IgM and IgG consistent with the timing of study and the time 144 course of antibody responses reported by others (7). We noted that there were 11 RT-PCR positive 145 samples that were negative by the lateral flow assay. As the study did not involve any independent 146 assessment of these samples by ELISA, it is assumed that these are false negative results. However, several 147 recent reports have raised the possibility of T-cell mediated immunity (8). In future studies, such 148 discrepancies must be addressed using alternate methods. 149

An analysis of results for the RT-PCR negative individuals revealed that 8 had antibodies to SARS CoV-2. 150

While the reasons for the negative RT-PCR results cannot be pinpointed, we speculate that the timing of 151 the test or suboptimal sample collection may have resulted in a false negative (9, 10). These results 152 suggest that employing a multiplicity of tests may be appropriate when diagnosing symptomatic 153 individuals. 154

The size of this study was small, n=99, with a 4:1 ratio of RT-PCR positive and negative individuals. The 155 small size allowed the study to be completed in a short period of time and the higher number of RT-PCR 156 positive individuals made it possible to probe the sensitivity of the assay in fuller detail. However, only a 157 limited evaluation of specificity was possible with this study design. At the time of initiating the study, this 158 was an acceptable trade-off as the manufacturer had claimed a specificity of 96% (oral communication) 159 and the need to verify sensitivity was more urgent. Unexpectedly, the number of true negative individuals 160 dropped from 20 to 12, greatly limiting the assessment of specificity. Future studies should include a larger 161 number of true negative samples, characterized by RT-PCR and an ELISA or Chemiluminescence 162 Immunoassay (CLIA) prior to enrollment in the study. 163

Overall, the data from this study support the use of the ImmunoQuick rapid test for serosurveillance. 164 165 All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this this version posted December 18, 2020. ; https://doi.org/10.1101/2020.12.16.20248303 doi: medRxiv preprint

SARS-CoV-2 diagnostic pipeline

Comparative assessment of multiple COVID-19 serological technologies supports continued 170 evaluation of point-of-care lateral flow assays in hospital and community healthcare settings

Evaluation of nine commercial SARS-CoV-2 immunoassays

Test 175 performance evaluation of SARS-CoV-2 serological assays

COVID-19: A report from the National COVID Scientific Advisory Panel. Wellcome Open Res

Detection of IgM and IgG antibodies in 181 patients with coronavirus disease 2019

Laboratory Diagnosis of COVID-19: Current Issues 187 and Challenges

Molecular diagnostic technologies for COVID-19: Limitations and challenges

Advanced Research

We thank Dr. N. Madhusudhana Rao from the Center for Cellular and Molecular Biology for helping us 192 with the ELISA tests on the venous samples. 193

The authors declare no conflict of interest. 195

This research did not receive any specific grant from funding agencies in the public, commercial, or not-197 for-profit sectors. 198