key: cord-296147-yfcp0xf2
authors: Mairesse, Antoine; Favresse, Julien; Eucher, Christine; Elsen, Marc; Tré-Hardy, Marie; Haventith, Caroline; Gruson, Damien; Dogné, Jean-Michel; Douxfils, Jonathan; Göbbels, Paul
title: High clinical performance and quantitative assessment of antibody kinetics using a dual recognition assay for the detection of SARS-CoV-2 IgM and IgG antibodies
date: 2020-08-25
journal: Clin Biochem
DOI: 10.1016/j.clinbiochem.2020.08.009
sha: 
doc_id: 296147
cord_uid: yfcp0xf2

Abstract Objectives Several serological SARS-CoV-2 immunoassays have been developed recently but require external validation before widespread use. This study aims at assessing the analytical and clinical performances of the iFlash® anti-SARS-CoV-2 chemiluminescence assay for the detection of both IgM and IgG antibodies. The kinetics of the antibody response was also evaluated. Design & Methods The precision, carry-over, linearity, limit of blank, detection and quantification were assessed. Sensitivity analysis was performed by using 178 sera collected from 154 RT-PCR confirmed patients COVID-19 samples. The specificity analysis was performed from 75 selected non-SARS-CoV-2 sera with a potential cross-reaction to the SARS-CoV-2 immunoassay. Results This iFlash® SARS-CoV-2 assay showed excellent analytical performances. After 2 weeks since symptom onset, the sensitivities for IgM and IgG were 62.2% (95% CI: 52.3-71.2%) and 92.9%% (95% CI: 85.7-96.7%), respectively by using the cut-off provided by the manufacturer. After cut-off optimization (i.e. >2.81 for IgM and >4.86 for IgG), the sensitivity for IgM and IgG were 81.6 (95% CI: 72.7-88.1%) and 95.9% (95% CI: 89.4-98.7%), respectively. Optimized cut-off for IgG improved the sensitivity to reach 100% (95%CI: 87.6-100) from 28 days since symptom onset. Conclusions This study shows that the iFlash® SARS-CoV-2 assay from YHLO biotechnology, has satisfactory analytical performances. Nevertheless, the sensitivity of the IgM is limited for a proper clinical use compared to IgG. The determination of anti-SARS-CoV-2 IgG antibodies from 28 days since symptom onset was associated with high sensitivity, especially using optimized cut-offs (i.e. 100%).

On December 30, 2019, the city of Wuhan, China, experienced an outbreak of unexplained pneumonia. On January 7, 2020, a new betacoronavirus, severe acute respiratory syndrome coronavirus (SARS-CoV-2) was identified. The number of confirmed cases worldwide exceeds 6.6 million and the number of deaths worldwide stands at 392,802 deaths (1, 2) . The gold standard method for the diagnosis of SARS-CoV-2 infection is (real-time) reverse transcription polymerase chain reaction (RT-PCR) in respiratory samples (3) . However, the accuracy of these molecular methods is variable and depends on several pre-analytical variables such as specimen collection, transport and storage (4) . In addition, clinical factors including time since infection and viral load also impact the sensitivity of the test (3).

Moreover, RT-PCR is not expected to detect past infection (5) and requires a high laboratory workload, skilled operators, valuable instruments and reagents, and crucial safety measures (6) . Detection of SARS-CoV-2 antibodies offers new perspectives. Serology has several roles including the identification of convalescent plasma donors, the screening of populations to determine exposure and immunity, and the diagnostic, especially in late-onset patients with a low viral load (7, 8) .

A wide range of serological immunoassays has been developed by in vitro diagnostic companies for the detection of SARS-CoV-2 antibodies, with different antigens target and formats (3, 6) . Independent validation by clinical laboratories are essential to access the analytical and clinical performances of these assays (3, 5-7, 9, 10) Consequently, the national Belgian authorities have planned a major validation campaign to assess the performances of these new-launched serological tests. The aim of this study was to evaluate the analytical and clinical performances of the iFlash ® SARS-CoV-2 antibodies (IgM and IgG) chemiluminescence assay (CLIA).

This retrospective study has been conducted from May 15 to 30, 2020 at the clinical biology laboratory of the Saint Nikolaus Hospital (Eupen, Belgium). A total of 178 serum samples coming from two hospitals (the Saint Nikolaus Hospital (Eupen, Belgium; n = 66), and Clinique St-Luc Bouge, Namur, Belgium; n = 112) were obtained from 154 patients confirmed positive to SARS-CoV-2 by RT-PCR and with COVID-19 symptoms. Antibody kinetics since the onset of symptoms was evaluated in the full cohort of patients for which the information on the onset of symptoms was available.

Blood samples collected from patients into serum tubes (BD Vacutainer ® 3.5 or 8.5 mL tubes, Becton Dickinson, New Jersey, USA), K 2 EDTA tubes (BD Vacutainer ® 4 mL tubes), or lithium-heparin plasma tubes (BD Vacutainer ® 4.0 mL tubes) according to standardized operating procedure and manufacturer's recommendations. Blood sampling was performed according to the recent guidelines (11). Two hundred thirty-five sera of patients taken during previous clinical requests for diagnostic purposes were collected from March 21 to May 30, 2020 and stored in the laboratory serum biobank at -20°C. Frozen samples were thawed one hour at room temperature on the day of the analysis. Re-thawed samples were vortexed before the analysis.

The iFlash ® anti-SARS-CoV-2 (YHLO biotechnology co., LTD, Shenzhen, China) is a chemiluminescent assay (CLIA) for the in vitro quantitative detection of IgM and IgG against the SARS-CoV-2 spike protein (S) and the nucleocapsid protein (N) in human serum and plasma. The analyzer converts a relative light unit (i.e. RLU) into an antibody titer (i.e. AU/mL) through a two-points calibration curve. According to the manufacturer, a titer ≥ 10AU/mL is considered positive (or reactive) for both IgM and IgG (12). The RT-PCR for SARS-CoV-2 determination in respiratory samples (nasopharyngeal swab samples) was performed on the cobas ® SARS-CoV-2 E-gene assay and ORF1abgene assay (Roche Diagnostics ® ).

Precision was evaluated by using two QC levels and three pools of human serum. Precision estimations were obtained by means of triplicates measurements of aliquots for a total of 5 consecutive days.

Repeatability (expressed as intra-run CV, %) and reproducibility (expressed as inter-run CV, %) of the method was calculated. Aliquots were stored at -20°C between analysis. Calculation was performed according to the Clinical and Laboratory Standards Institute (CLSI) EP15-A3 protocol (13).

The diluent provided by the manufacturer was used as blank sample to determine the limit of blank (LOB), detection (LOD) and quantification (LOQ).

The LOB has been determined by running the blank sample on three separate occasion to verify that the results are well <10 AU/mL. The LOD and LOQ have been determined by running 30 analyses of the blank sample using the following equations according the SH GTA 04 document -revision 1 of the COFRAC (14) .

-LOD = mean of the 30 measurements + 3*standard deviation -LOQ = mean of the 30 measurements + 10*standard deviation  Linearity assessment Linearity was evaluated according to CLSI EP-06. A sample with a high total antibody levels (i.e. for IgM: 334.64 AU/mL; for IgG: 71.57 AU/mL) was analyzed and diluted by a factor 2 on 5 consecutive dilutions. The manufacturer's diluent was used for the dilution. Observed values were compared to the expected ones and polynomial regression was calculated.

A sample with high total antibody levels (i.e. for IgM: 124.02; for IgG: 125.67) was analyzed in triplicate (A1, A2, A3) and followed by a negative sample (i.e. for IgM: 0.49; for IgG: 0.35) also analyzed in triplicate (B1, B2, B3). The carry-over is calculated using this following equation: (B1-B3)/(A3-B3) x 100. A carry-over below 1% is considered satisfactory. Samples were subdivided according to different categories since symptom onset as follow: 0-6 days: 45 sera; 7-13 days: 35 sera; 14-20 days: 37 sera; 21-27 days: 29 sera; 28 days or more: 32 sera. Clinical sensitivity for SARS-Cov-2 serological test depending on the onset of COVID-19 symptoms was carried out with the manufacturer's cut-off (>10 AU/mL for both IgM and IgG) and with ROC curve adapted cut-offs (2.81 AU/mL for IgM; 4.86 AU/mL for IgG).

Antibody kinetics since symptom onset was evaluated using the following time frames (d, days): 0-2 d, 

Descriptive statistics were used to analyze the data. Sensitivity was defined as the proportion of correctly identified COVID-19 positive patients since symptom onset. Specificity was defined as the proportion of naïve patients or healthy volunteers classified as negative. The ROC area under the curve (AUC) was calculated as the fraction of true positive and false positive determined according to the manufacturer's cut-off values for positive results. Samples included for ROC curves analyses were sera obtained from at least two weeks after symptoms onset (n = 98), sera selected to assess cross-reactivity (n = 37) and sera from healthy volunteers (n = 38). Data analysis was performed using XLSTAT ® software (version 2019.2.2, Addinsoft, Paris, France). P value < 0.05 was used as a significance level. Our study fulfilled the Ethical principles provided by the Declaration of Helsinki.

Repeatability and reproducibility results are summarized in Table 1 . Repeatability and reproducibility were ≤7.5% and ≤13.3%, respectively.

The limit of blank, detection and quantification were 0.15 AU/mL and 0.035 AU/mL; 0.242 AU/mL and 0.441 for IgM; 0.052 AU/mL and 0.097 AU/mL for IgG, respectively. 

The calculated specificity was 98.7% (95% CI: 92.0-100%) for IgM and 100% (95% CI: 94.0-100%) for IgG by using the manufacturer's cut-off (i.e. ≥10 AU/mL). Using optimized cut-off, specificity was 94.7% (95% CI: 86.6-98.2%) for IgM and 100% for IgG (95% CI: 94.0-100%) ( Table 2 and Supplementary Figure 2&3) .

Sensitivities for IgM and IgG according to different time categories since symptom onset are represented in Table 2 . Before 14 days since symptom onset, sensitivities were not sufficient to be reliable in clinical practice. After 2 weeks since symptom onset, the sensitivity for IgM was 62.2% (95% CI: 52.3-71.2%) and was 92.9%% (95% CI: 85.7-96.7%) for IgG by using the cut-off provided by the manufacturer.

Using the optimized cut-off, the sensibility for IgM and for IgG (i.e. >2.81 for IgM and >4.86 for IgG) were 81.6 (95% CI: 72.7-88.1%) and 95.9% (95% CI: 89.4-98.7%), respectively. From 28 days since symptoms onset, the sensitivity for IgG increased to 100 % by using the optimized cut-off ( Table 2 and Supplementary Figure 2&3 ).

The Figure 1 shows the IgG kinetics at different days from symptom onset. The IgG kinetics gradually increased from day 0 to day 62, to achieve highest IgG concentrations between days 31 and 40. The IgM kinetics increased from day 0 to day 17 and then gradually decrease from day 18 to day 62. Standard deviations of the mean for IgM were very large arguing for a higher inter-individual variation compared to IgG.

Serological testing is a useful strategy for the diagnosis, the characterization of the course of the disease, for identifying convalescent plasma donors as well as for epidemiological study, lockdown exit programs and COVID-19 vaccine development [5, 6, 8, 16] . Being rapidly developed and flooding the market, these tests need to be evaluated both in terms of analytical and clinical performances (6, 7, 9, 15) . The aims of the present study were to evaluate the analytical and clinical performances of the iFlash ® anti-SARS-CoV-2 CLIA assay for IgM and IgG antibodies with a large cohort of COVID-19 patients, to provide an external validation of this test and to evaluate the antibody kinetics since symptom onset.

Repeatability and reproducibility studies had CVs below 7.5% and 11.9% for IgM, and below 5.8% and 13.3% for IgG. Near the optimized cut-offs, CVs for repeatability and reproducibility were 1.1% and 5.7% for IgM, and 1.8% and 2.9% for IgG. These results are consistent with those published recently (16) . The carry-over was negligible for IgM and IgG antibodies (<0.01%). We found a LOQ of 0.441 AU/mL for IgM and 0.097 AU/mL for IgG, which is below the value we found for the optimized cutoffs (i.e. IgM: 2.81 AU/mL and IgG: 4.86 AU/mL). The linearity of the dilutions has been performed in order to assess the ability of the method to provide direct results proportional to the concentration of IgM and IgG in the test product samples. The results obtained showed excellent linearity from ±330 AU/mL to ±5 AU/mL for IgM and from ±70 AU/mL to ± 2 AU/mL for IgG. Linearity appears poor at the highest values for IgG as already described on another platform (data not shown) (17) .

A total of 253 sera were included to evaluate the clinical performances of the assay in our study. IgG (16) . These were in line with our observations. The sensitivity of the iFlash ® SARS-CoV-2 assay was also evaluated in samples collected at various times since symptom onset. In our investigation the sensitivity of the test for IgG increased progressively to reach a sensitivity of 100% (95%CI: 87.6-100%) after 28 days by using an optimized cut-off (i.e. >4.86 AU/mL). Using the optimized cut-off (i.e. >2.81 AU/mL), the IgM sensitivity reached its higher level (86.5%) between 14 and 20 days since symptom onset. Nevertheless, it remained globally of less clinical interest compared to IgG. The high sensitivity observed with the iFlash ® SARS-CoV-2 assay may be explained by its capacity to recognize both the antibodies directed against the spike proteins and the nucleocapside.

Using ROC curve adapted cut-offs (10.0 AU/mL for IgM and 7.1 AU/mL for IgG), Infantino et al. found sensitivities and specificities of 73.3% and 92.2%, and of 83.3% and 100%, for IgM and IgG, respectively (18). However, authors only included 61 COVID-19 patients, did not subdivide patients according to days from symptom onset, and selected patients with a short timeframe inclusion, i.e. from 8 to 17 days only. In our study, we found that the sensitivity increased with time from symptom onset to reach its higher level from 28 days. The suboptimal sensitivities they observed is highly due to a selection bias. Plebani et al. only included samples collected after 11 days from the onset of symptoms, with a mean time interval from symptom of 24 days (standard deviation ±11 days; range 12 -54 days) (16) . They did not categorized patients according to days from symptom onset, and only included a limited number of COVID-19 patients, i.e. 64 SARS-CoV-2 patients. In this study, the use of a redefined cut-off slightly increases the sensitivity of the assay, from 92.9% to 95.9%.

Existing data support that seroconversion occurs approximately 7 to 14 days after the onset of symptoms (8, (19) (20) (21) (22) . In our study, the kinetics of IgM and IgG was evaluated during a long-term period (2 months) at regular time intervals (every 3 days) and on a significant cohort. Anti-SARS-CoV-2 IgG antibodies gradually increased since the onset of COVID-19 symptoms to achieve a 100% sensitivity from days 28 ( Table 2) . After 28 days, we observed a plateau phase and we can assume aslight decrease in IgG levels. This finding is in accordance with recently published data (23).

Further designed long term studies are needed to evaluate IgG kinetics and especially the persistence of potentially neutralizing antibodies. Recently there has been some concern about antibody kinetics and the persistence of these antibodies especially in asymptomatic patients.

They could only be present for 2 to 3 months (24) . In this present study we only had patients with symptoms for 2 months maximum.

Before days 14, the improved sensitivity observed when using the optimized cut-off for IgM compared to IgG was only modest ( Table 2) . Padoan et al. have also studied the kinetics of IgM and IgG on a smaller cohort (i.e. on 37 COVID-19 patients) and observed a 100% sensitivity for IgG 12 days after the onset of symptoms while a lower sensitivity (88%) was reported for IgM (25), and further confirmed in another study (26). Our investigation confirms the suboptimal sensitivities reported for IgM on a larger study population and using another analyzer strengthening the hypothesis that IgM measurement may be less relevant for assessing the seroprevalence in previously exposed patients.

In conclusion, this study shows that the iFlash ® SARS-CoV-2 assay from YHLO biotechnology, has satisfactory analytical performances. The determination of anti-SARS-CoV-2 IgG antibodies from 28 days since symptom onset was associated to high sensitivity, especially using an optimized cut-off, which strengthens the strategy of cut-off optimization for these SARS-CoV-2 serological assays.

Among the authors, Jonathan Douxfils is chief executive officer and founder of QUALIblood sa and reports personal fees from Diagnostica Stago, Roche, Roche Diagnostics, Daiichi-Sankyo, and Portola, outside the submitted work. The other authors do not have conflict of interests. Table 1 : Mean (AU/mL), repeatability (CV %) and reproducibility (CV %) assessed on three patient pool and two QC levels. Table 2 : Clinical performances of iFlash ® CLIA analyzer for anti-SARS CoV-2 IgM and IgG since symptom onset with the manufacturer's cut-off and with optimized cut-offs. 

Covid-19 -Navigating the Uncharted. The New England journal of medicine

Vitro Diagnostic Assays for COVID-19: Recent Advances and Emerging Trends

Potential preanalytical and analytical vulnerabilities in the laboratory diagnosis of coronavirus disease 2019 (COVID-19). Clinical chemistry and laboratory medicine : CCLM / FESCC

The important role of serology for COVID-19 control

Molecular, serological, and biochemical diagnosis and monitoring of COVID-19: IFCC taskforce evaluation of the latest evidence. Clinical chemistry and laboratory medicine : CCLM / FESCC

SARS-CoV-2 Serology: Much Hype, Little Data. Clinical chemistry

Antibody responses to SARS

CoV-2 in patients with COVID-19. Nature medicine

Antibody tests for COVID-19: drawing attention to the importance of analytical specificity

Validation of a chemiluminescent assay for specific SARS-CoV-2 antibody. Clinical chemistry and laboratory medicine : CCLM / FESCC

Joint EFLM-COLABIOCLI Recommendation for venous blood sampling. Clinical chemistry and laboratory medicine : CCLM / FESCC

iFlash Anti-SARS-CoV-2 (2019-nCoV IgG)

Clinical and Laboratory Standards Institute. EP15A3 -User verification of precision and estimation of bias; approved guideline

Guide Technique d'Accréditation de Vérification (Portée A) / Validation (Portée B) des

Méthodes en Biologie Médicale -Document SH GTA 04 (révision 01)

COVID-19 and Postinfection Immunity: Limited Evidence, Many Remaining Questions

Diagnostic performances and thresholds: the key to harmonization in serological SARS-CoV-2 assays?

Analytical performances of a chemiluminescence immunoassay for SARS-CoV-2 IgM/IgG and antibody kinetics. Clinical chemistry and laboratory medicine : CCLM / FESCC

Diagnostic accuracy of an automated chemiluminescent immunoassay for anti-SARS-CoV-2 IgM and IgG antibodies: an Italian experience

Molecular, serological, and biochemical diagnosis and monitoring of COVID-19: IFCC taskforce evaluation of the latest evidence. Clinical chemistry and laboratory medicine : CCLM / FESCC

Severe Acute Respiratory Syndrome Coronavirus 2-Specific Antibody Responses in Coronavirus Disease

Emerging infectious diseases

Elecsys electrochemiluminescent immunoassay for the detection of SARS-CoV-2 total antibodies

The role of serology for COVID-19 control: Population, kinetics and test performance do matter

Response of anti-SARS-CoV-2 total antibodies to nucleocapsid antigen in COVID-19 patients: a longitudinal study

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

Analytical performances of a chemiluminescence immunoassay for SARS-CoV-2 IgM/IgG and antibody kinetics. Clinical chemistry and laboratory medicine : CCLM / FESCC

Evaluation of two automated and three rapid lateral flow immunoassays for the detection of