A Dried Blood Spot protocol for high-throughput quantitative analysis of SARS-CoV-2 RBD serology based on the Roche Elecsys system

Abstract

SARS-CoV-2 spreads pandemically since 2020; in 2021, effective vaccinations became available and vaccination campaigns commenced. Still, it is hard to track the spread of the infection or to assess vaccination success in the broader population. Measuring specific anti-SARS-CoV-2 antibodies is the most effective tool to track the spread of the infection or successful vaccinations. The need for venous-blood sampling however poses a significant barrier for large studies. Dried-blood-spots on filter-cards (DBS) have been used for SARS-CoV-2 serology in our laboratory, but so far not to follow quantitative SARS-CoV-2 anti-spike reactivity in a longitudinal cohort. We developed a semi-automated protocol or quantitative SARS-CoV-2 anti-spike serology from self-sampled DBS, validating it in a cohort of matched DBS and venous-blood samples (n = 825). We investigated chromatographic effects, reproducibility, and carry-over effects and calculated a positivity threshold as well as a conversion formula to determine the quantitative binding units in the DBS with confidence intervals. Sensitivity and specificity reached 96.63% and 97.81%, respectively, compared to the same test performed in paired venous samples. Between a signal of 0.018 and 250 U/mL, we calculated a correction formula. Measuring longitudinal samples during vaccinations, we demonstrated relative changes in titers over time in several individuals and in a longitudinal cohort over four follow-ups. DBS sampling has proven itself for anti-nucleocapsid serosurveys in our laboratory. Similarly, anti-spike high-throughput DBS serology is feasible as a complementary assay. Quantitative measurements are accurate enough to follow titer dynamics in populations also after vaccination campaigns. This work was supported by the Bavarian State Ministry of Science and the Arts; LMU University Hospital, LMU Munich; Helmholtz Center Munich; University of Bonn; University of Bielefeld; German Ministry for Education and Research (proj. nr.: 01KI20271 and others) and the Medical Biodefense Research Program of the Bundeswehr Medical Service. Roche Diagnostics provided kits and machines for analyses at discounted rates. The project is funded also by the European-wide Consortium ORCHESTRA. The ORCHESTRA project has received funding from the European Union's Horizon 2020 research and innovation program under grant agreement No 101016167. The views expressed in this publication are the sole responsibility of the author, and the Commission is not responsible for any use that may be made of the information it contains.IMPORTANCESARS-CoV-2 has been spreading globally as a pandemic since 2020. To determine the prevalence of SARS-CoV-2 antibodies among populations, the most effective public health tool is measuring specific anti-SARS-CoV-2 antibodies induced by infection or vaccination. However, conducting large-scale studies that involve venous-blood sampling is challenging due to the associated feasibility and cost issues. A more cost-efficient and less invasive method for SARS-CoV-2 serological testing is using Dried-Blood-Spots on filter cards (DBS). In this paper, we have developed a semi-automated protocol for quantifying SARS-CoV-2 anti-spike antibodies from self-collected DBS. Our laboratory has previously successfully used DBS sampling for anti-nucleocapsid antibody surveys. Likewise, conducting high-throughput DBS serology for anti-spike antibodies is feasible as an additional test that can be performed using the same sample preparation as the anti-nucleocapsid analysis. The quantitative measurements obtained are accurate enough to track the dynamics of antibody levels in populations, even after vaccination campaigns.

Keywords: COVID-19; DBS; Roche Elecsys; SARS-CoV-2; Spike; antibody; quantitative; serology.

Fig 1

Description of the data. DBS eluates from patients of our in-house KoCo19 cohort and volunteer donors (n = 825) are used. The dashed vertical line denotes the empirically determined cutoff value (0.115) for result classification. (A) Frequency distribution of detected antibody levels against SARS-CoV-2 spike/RBD in vaccinated or naturally infected individuals. The insert in the top right represents a zoom-in on the y-axis to allow visualization of the cutoff region. (B) Empirical cutoff determination (dashed vertical line). Percentages of false positives/negatives depending on a variable threshold for DBS are shown in red or black, respectively. The cutoff was chosen to minimize false positive results.

Fig 2

Scatterplot illustrating the linear relation on log10–log10 scale between antibody units detected in the DBS eluate (x-axis) and the corresponding venous plasma (y-axis) (n = 825). Correlation is calculated using the Spearman method. The dashed horizontal line denotes the manufacturer’s cutoff value for plasma result classification (0.8), while the dashed vertical line is the empirically determined cutoff value for DBS result classification (0.115). The solid black line represents a linear regression, while the dashed line is the LOESS (locally estimated scatterplot smoothing or local regression) modeling different types of associations. The gray region is the 95% CI of the linear/LOESS estimate.

Fig 3

Bland-Altman plot illustrating the differences as a function of means for discrepancies between plasma-values and DBS-values converted to plasma units. Both plasma and converted DBS values are displayed as log10 scales for the analysis. The mean value signal difference is ${3·39\cdot 10}^{-15}$ BAU with 95% of the values in the interval (−0.588 BAU; +0.588 BAU).

Fig 4

Temporal evolution of the anti-S titers in BNT162b2 vaccinated SARS-CoV-2 naïve subjects. The antibody response is shown in back-calculated venous sample BAU/mL. Vertical dashed lines indicate the date of the second vaccination, set to day 0. (A) Time series of two subjects over 150 days (Volunteers 1 and 5). (B) Titer-increase after the second dose in three other subjects (Volunteers 2, 3, and 4).

Fig 5

Distribution of DBS eluate values for six different repetitively punched DBS samples. Blood spots were punched on the drying edge (coffee stain middle boxplot) or the center (volcano effect right boxplot) of each respective blood spot. The same samples were also punched using the automatic system of the Panthera machines, aleatory (aleatory, left boxplot). For all the samples, except 99000202, significant differences could be detected by applying the Kruskal-Wallis test (P < 0·001), demonstrating higher values at the edges of the blood drop, which would be consistent with a slight coffee stain effect.

Fig 6

Scatterplot of four follow-ups of the KoCo19 cohort for people who participated in all rounds (n = 3,040). The Ro-N-Ig measurement from DBS is abbreviated with “N,” and Ro-RBD-Ig-quant from the same DBS is abbreviated “S.” Positivity is represented with “+,” negative, below cutoff with “−.” The color code is defined by the status of the respective subject in the (A) second, (B) third, and (C) fourth follow-up, respectively (represented by “FU”). Blue dots represent N−S−, orange dots represent N + S + , gray spots are N−S + and pink dots are N + S−, considering the left column as reference for color-coding. Samples above the nonlinear range of Ro-RBD-Ig-quant (solid black line at 9730.4 for back calculated plasma BAU/mL) were not diluted. (A) Evolution from second to third follow-up. Left: Second follow-up sampled between March and April 2021; right: Third follow-up sampled between July and September 2021. (B) Evolution from third to fourth follow-up, sampled between October and December 2021 and (C) evolution from the fourth to the fifth follow-up, sampled between May and July 2022.